This is kind of like finding a business with excess cash. While it's true that a business can have too much cash from an efficiency point of view, finding more cash on the balance sheet than you expected is always a good thing, right? The point in each case is that the headline numbers (EPS, P/E, etc.) sometimes lie - and an inordinate number of bargains are found where such "lies" exist – simply, because others aren’t looking there (it’s a less conspicuous bargain).
But wouldn't that mean the company wasn't reinvesting in P&E?
Some businesses have a very strong relationship between the value of the assets in the business and earnings.
Others have almost no correlation between the two. For an example of a business that will likely have very different ROAs from year to year (and longer-term), look at Forward Industries (FORD). A less extreme example is Craftmade International (OTC:CRFT), and further down the spectrum (but still very asset light), you have companies like Timberland (TBL) and K-Swiss (KSWS).
For an example of a business, that long-term at least, has to add to assets to add to earnings look at Village Supermarket (VLGEA). In this case (as in the case of most retailers), the long-term correlation between assets and earnings is somewhat obscured by operating leverage. However, in VLGEA's case (logically at least), you do recognize that a supermarket’s earnings will be determined in large part by the number (and size) of the stores being operated.
Also on this side of the spectrum (businesses with a strong long-term correlation between assets and earnings) you have various businesses that own distinct, identifiable assets such as: theme parks, pipelines, parking lots, bowling alleys, golf courses, hotels, etc. Of course, you also have asset-heavy manufacturing businesses, especially in price sensitive, commodity-like products.
Both of these types of businesses tend to have more predictable returns on assets (at least on the margins). I add the qualifier, because it’s a rare business that is both capital intensive and highly profitable - although I'm sure you could name a handful of such conglomerates.
Some asset-light businesses have predictable returns on assets – not so much because there is a strong correlation between assets and earnings, but rather because there is the absence of disruptive change and some real protection from price competition. An example from my recent podcast on this topic would be McCormick (MKC) – a business that has a fairly predictable ROA largely because it’s simply a great business (albeit a slow growth business).
One of the greatest investing conundrums is the fact that it is usually easiest to reinvest retained earnings at past rates of return in a poor business and hardest to reinvest retained earnings at past rates of return in a good business.
In other words, many of the least limited businesses tend to be the least profitable, and many of the most profitable tend to be the most limited. That’s why you hear me talk so much about 'franchises' and 'niches'.
If I were forced to invest every dime I had in a single business and hold it for the rest of my life, the first characteristic I would look for is a business with virtually no need for maintenance cap-ex.
The 'Pleasant Surprise' Effect
The 'pleasant surprise' is finding that the GAAP earnings are lower than the actual amount of cash a 100% owner would be able to extract from the business, if he chose not to expand it (via additional spending).
A lot of companies have depreciation charges that adequately mirror maintenance cap-ex requirements. That isn’t to say the two items are necessarily the same amount; but, the extent to which they diverge from each other is not terribly specific to the business. The most obvious reason for a major divergence is inflation. Regardless, stocks with similar P/E ratios generally also have similar 'owners’ earnings' multiples.
This isn’t true if the assets on the book don’t really need to be replaced to maintain the same earnings power. Some businesses do have assets that need to be maintained (brand, technology, etc.) – but, these assets are maintained as a part of daily operations and are not broken out as a separate item (it would be nearly impossible to separate “brand maintenance” from other expenses anyway).
The most conspicuous examples of such brand maintenance are all the ads you see for GEICO, PetMeds Express (PETS), etc. At least in these two cases, there is no doubt such advertising creates an economic asset that helps generate earnings in future periods.
Such spending is not treated as a capital investment. Therefore, GAAP accounting tends to exaggerate the actual cost of day-to-day operations for these businesses and understate the amount of additional investment in the business (both GEICO and PetMeds Express are heavily investing in future growth – it’s just that those investments aren’t in the form of tangible assets such as a new production plant).
I’m sure it sounds like I’m taking quite a leap here. After all, there have been businesses that argued for the amortization of certain operating expenses that clearly did not have much of a useful life. You may remember a few such instances from the late 90s. However, a review of the past financials for PetMeds Express illustrates my point. Since 2000, the company’s revenues have increased roughly tenfold while net Property, Plant, and Equipment has been cut by two-thirds.
The reason? Advertising. The majority of the company’s operating expenses are advertising expenses. Let me put the difference between the intangible asset of the PetMeds brand and all of the company’s tangible assets into perspective.
In 2005, depreciation expenses totaled less than 0.5% of sales while advertising expenses totaled more than 15% of sales. In previous years, advertising expenses were even greater as a percentage of sales.
My point is simply that some of this advertising spending (and I’m guessing a whole lot) creates economic benefits in future periods. In other words, economically, part of that advertising spending is an investment, not an expense. I’m not saying GAAP accounting should treat the advertising as an investment in an intangible asset, I’m just saying, the advertising is such an investment.
So, the pleasant surprise is the phantom investment. GAAP earnings in previous years were lower than economic earnings, because an investment in future growth was treated as an operating expense.
Again, I think this is, in fact, how the item should be treated by accountants. However, investors need to recognize the distinction and adjust their expectations accordingly.
To better explain what all this talk of accounting for advertising is about, I’ll provide an excerpt from the company’s 10-K:
The Company's advertising expense consists primarily of television advertising, internet marketing, and direct mail/print advertising. Television costs are expensed as the advertisements are televised. Internet costs are expensed in the month incurred and direct mail/print advertising costs are expensed when the related catalog and postcards are produced, distributed or superseded.
Simply put, the hit to earnings is immediate, while the full economic benefits are only realized over a period of many years.
That’s what I meant when I wrote that the EPS number (and thus the P/E ratio) “sometimes lie.” This is one of those times. An owner would see the advertising spending differently than the GAAP portrayal. Therefore, he would believe the true P/E ratio was lower than it appeared to be.
I hope this makes sense.
The Value of Intangibles
Intangible assets are often harder to reproduce than tangible assets.
There is a nearly infinite potential supply of new plants and stores if a competitor wants to build them - and they can usually be built at the same cost regardless of who builds them.
Already, if a competitor wanted to reproduce the PetMeds or GEICO brands, they would have to spend considerably more than those companies did, because both brands are fairly entrenched within our minds - they've staked a claim to the territory in our mind where we think 'pet medicines' or 'car insurance' (and that annoying talking lizard).
You can't reproduce those brands at the same cost. Furthermore, in both of these cases, you'd have to lose money or accept a much narrower margin while you did build the brand up. So, while the barriers to entry may not be obvious, the barriers to profitability and dominance are quite clear.
Both companies already own a little piece of your mind. That’s valuable real estate – even if it doesn’t show up on the books.
So how does one parse the numbers to find these hidden bargains?
There is no purely quantitative way of doing this. Qualitative considerations loom large in any estimate of cap-ex requirements, because the nature of the business and the competitive position of the firm are key determinants of how effective new cap-ex spending is.
If you can't explain why one company spends less on cap-ex than its competitors, you have to assume the current skimping on cap-ex is not sustainable.
One important caveat though: many companies in the same industry are not competitors, and therefore cap-ex comparisons between them are of little use. For example, Strattec (STRT) and Lear (LEA) both make auto parts. However, they aren't competitors. Lear makes interiors; Strattec makes locks.
The lock business is not the same as the interior business. The industries aren't equally profitable and they aren't equally competitive. You have to analyze each business separately - just as you can't lump Amazon.com (AMZN) and PetMed Express together, even though they both sell a lot of stuff on the web.
Any consideration of cap-ex spending and how it's really divided between 'maintenance' and 'investment' has to begin with your assessment of the nature of the industry in general and the specific competitive position of the company you're looking at.
Then, you can start making cap-ex comparisons. But, don't allow yourself to become unduly wed to the numbers. Bring your understanding of what's needed to maintain and expand the particular business and what competitors are likely to do (and the unintended consequences those likely actions will produce).
Some industries are easy. Unless you have a very special case, a steel company's cap-ex will be determined by the long-term economics of the steel industry (which is not extraordinarily profitable). You aren't going to find one company that can skimp on capital spending - they all have to ante up each round.
At any one time, the numbers for the last few years may not make this fact obvious, but you'll know it, because of the qualitative judgments you bring to your analysis of any particular steel company. Just as your qualitative judgments regarding PetMed Express would have helped you realize the low cap-ex spending there was perfectly fine, because the real investment was the advertising. These are the things the numbers alone can’t tell you.
For a related article on this topic, check out my April 25, 2006 post titled On Inflexible Enterprises.