June 13, 2005
Google, iPods and George Foreman Grills
I'm going to immediately preface this update with an essential fact of investing: value is not a timing tool. To identify a security as overvalued does not imply a decline in price over the near term, or even over a short period of years. Rather, overvaluation implies only an unsatisfactory tradeoff between the current price and future cash flows. Overvaluation implies that given the current security price, the stream of future cash flows delivered by that security is likely to result in an unsatisfactory long term investment return. Similarly, “fair value” is not a target price. Rather, it's a price that would most probably result in reasonable long-term investment returns, given the likely growth and risk of the future cash flows being purchased. We like to think that markets tend to move toward “fair value” over the long-term, but it's essential to remember that this “mean reversion” can take 5 years, 10 years or longer.
If you spend any time at all studying valuation models, you'll quickly discover that you can justify nearly any price you want for a security, provided you're willing to assume fast enough growth and low enough long-term returns. In order to transform these models from algebraic curiosities into useful investment tools, the crucial step is to make sure – and this is essential – that your assumptions are reasonable. Reasonable assumptions involve more than saying, well, yeah, 50% annual growth is too high, so I'll bring it down to 30%. Reasonable assumptions mean that you've examined the actual long-term growth and investment returns of hundreds or thousands of companies, and you've made some attempt to understand the business models that produce different growth patterns – why most explosive growth rates suddenly flame out, or why one moderately growing company can keep on growing, while another with a very similar growth pattern just drops off the chart.
That's much of what I've worked on since about 1981, so I'll share one of my main insights: earnings just don't grow that fast over the long-term.
That's certainly true for the market as a whole, where I've frequently pointed out that S&P 500 earnings growth, measured from peak-to-peak across economic cycles, has never significantly exceeded 6% annually. The S&P prices its current, peak earnings at a multiple of 20, which is very high historically. Unfortunately, that multiple probably understates the level of valuations, because earnings remain very elevated relative to other fundamentals like revenues, dividends, and book values, because of unusually wide profit margins. Profit margins, however, are cyclical and mean-reverting. As Jeremy Grantham recently noted “If profit margins aren't mean-reverting, then capitalism is broken.”
Importantly, the observation that earnings just don't grow that fast is also true for individual stocks. It's certainly possible for companies to grow at explosive rates during the emergence of a new technology, or during the period where a leading company grows to the saturation point in its industry. But except for explosive growth occurring off of a low base, you'll find very few stocks with established revenues and earnings for which a growth rate of even 15% annually is appropriate for more than about 7-10 years, and even shorter horizons for higher growth rates.
Consider these comments from Warren Buffett, which appeared in Berkshire Hathaway's 2000 annual report: “Here's a test: Examine the record of, say, the 200 highest earning companies from 1970 or 1980 and tabulate how many have increased per-share earnings by 15% annually since those dates. You will find that only a handful have. I would wager you a very significant sum that fewer than 10 of the 200 most profitable companies in 2000 will attain 15% annual growth in earnings-per-share over the next 20 years.”
Bandwagons and Saturation
Often, when we see explosive growth in a particular company, we're really seeing a bandwagon effect in which a large number of customers adopt the product at the same time. Investors often extrapolate that sort of “bandwagon spike” as if it was sustainable growth, where in fact it is often the crescendo of adoption for that particular product.
There were countless examples of this in the late 1990's tech boom. By some software industry estimates, nearly 40% of the software purchased by companies in the late-1990's was, in fact, never installed. In a period where the announcement of a website was enough to give even tire companies a jump in their stock price, the frenzy to become networked led to an enormous amount of overinvestment. If you page through charts and financials of most leading tech companies at the time, you'll see that their revenues and earnings exploded higher, and then abruptly crashed or leveled off.
This isn't just a tech phenomenon either. Several years ago, Salton (SFP) enjoyed huge sales of the George Foreman grill, and investors drove the stock from just $1 a share in 1995 to nearly $60 a few years later. The explosive growth in the stock price was the result of that same confusion between a bandwagon spike and sustainable growth. The George Foreman grill was both novel and useful, and its simultaneous adoption led to an explosive crescendo that investors confused for something sustainable. Once that spike occurred, the market was essentially saturated. Despite growth in new users of the grill, the company could never achieve the sales it enjoyed during that period of saturation. Last week, Salton closed at $1.40 a share, despite the fact that the George Foreman grills are still great products and still enjoy good sales – I use mine all the time.
My suspicion is that investors are responding to the simultaneous adoption of Apple's iPod in exactly the same way. Hey, it's a neat product, and a very creative company, but investors who confuse simultaneous adoption for sustainable growth are going to have their hats handed to them. Despite Apple's great creative success over the years, it hasn't prevented the long-term stock chart from looking like a panoramic shot of the Appalachian mountains .
The chart below shows the difference between three basic types of growth. The blue line depicts exponential growth, which produces ever-increasing values over time. This is the type of growth that analysts often implicitly assume when they make estimates of future earnings and revenues. Unfortunately, it's not at all typical of corporate earnings growth, and is extremely rare for high growth rates. The red line depicts “logistic growth,” which is the sort of growth you see in markets and other systems that reach a saturation point. Population growth in a closed environment, for example, is often modeled as a logistic process. This sort of growth profile is much more common among corporations – adoption of the product up to a certain (possibly large) market share, followed by continued business but limited growth. The green line depicts what might be called an “extraction” process – that is, you extract your new business primarily from the pool of customers you aren't already servicing. This is the sort of pattern you typically see in products that don't involve continual growth in customer expenditure or sustained repeat business from a large proportion of customers. The George Foreman grill fits nicely into this category, as I suspect Apple's iPod will.
Look at the early portion of each graph. It's important to notice that you can't really identify the true growth process by looking at the early growth of a company. You have to actually think, read, look at the products, stare at the business model, compare it with other situations you've observed, and do all sorts of things that you can't squeeze out of equations or raw numbers. It's easy to value any stock at any price you want if you're willing to plug numbers into a spreadsheet and ignore the facts.
1/5 of General Electric?
[Post-note: In response to questions about the calculations below, the figures in this article refer to the expected discounted value of operating cash flows that will be delivered to investors by the business. That is, the figures assume a 100% purchase of the operating assets. The distinction is this - if a company issues stock and just holds the proceeds in cash, the value of the company is the value of the operating business plus the value of the cash proceeds from the stock sale. Trivially, if you invest $1 billion in a lemonade stand, the proceeds can be deposited in the bank and the business will be worth $1 billion, even if nobody ever makes lemonade. The lemonade stand itself, based on operating cash flows as a standalone entity, might be worth only $24, which you would add to the "cash and marketable securities" to get the value of the company. This is important because Google has been aggressive in issuing additional shares of stock since its IPO, so the per-share value of issuance proceeds invested in "cash and marketable securities" reached about $60 in 2007. Note also that these comments imply nothing about near-term price direction, only about disappointing long-term returns that investors will probably earn by purchasing the stock a price far greater than those stock issuance proceeds plus a conservative amount for the operating business itself.]
Which brings us to Google. Initially, I estimated Google to be worth about $24 a share. It has since enjoyed some very good operating surprises. I'd currently estimate its value somewhere in the $30's
No zero is missing in that last sentence, though the quickest way for the company to substantially enhance its intrinsic value would be to buy everything it possibly can with its own overvalued currency. Having made similar comments regarding the value of Cisco, Sun, EMC and Oracle near the peak of the tech bubble, with value estimates (which turned out to be slightly optimistic) at small fractions of their going prices, I'm pretty comfortable with that figure.
The difficulty with Google isn't in the product. It's neat. It's hip. I use it almost daily. The real question is this – why do we naturally assume that Google's revenues and earnings are going to grow exponentially from here? In a competitive market, with few barriers to entry and no particular brand loyalty, an advertising company that's valued at 1/5 the market cap of General Electric is not likely to mount a secure defense for its revenue stream. Unlike Microsoft, Adobe, or even Yahoo and Ebay, there's currently no benefit to users in aggregating around the same product as their “standard” (note to Google – this is really where you ought to be focusing your attention) and no high-cost obstacle to entry aside from smart statistical and computing algorithms. Therefore, there's no natural monopoly that would lead to a defensible competitive advantage. Sure, these guys are smart, and deliver useful, consumer oriented products. But when you value a company at 20 times revenues and over 100 times earnings, you're going to invite competition from some very, very intelligent people.
Given the company's business model, the long-term growth process is much more likely to represent a blend of logistic and extraction processes than the explosive exponential process that analysts seem to be taking for granted. And wow, will that make a difference over time.
My impression is that the explosive growth we've seen in Google's revenues recently has been very much a “simultaneous adoption” effect. With the enormous notoriety the company has enjoyed, it's natural that advertisers would want to try out that venue, and even pay a premium to do it. At present, however, the value of a click from Google is worth no more to an advertiser than a click from another source, so even if the value of Google's free product (search) is good, the value of their paid product (advertising clicks) is no higher than its competitors. And nobody, free or paid, has any inherent loyalty or reason to stay if another product emerges that is marginally better. This isn't the stuff that defensible profits are made of.
What investors seem to be doing is paying an awful, awful lot for future creativity. That might be reasonable, to at least some extent, if the company's dominant position in the industry had characteristics of a natural monopoly. But it doesn't. Usefulness of a product is insufficient for profitability unless it is coupled with scarcity. Why does Wall Street take it as given that the company will grow explosively into the indefinite future, much less maintain its profitability or market share over time?
To paraphrase Grantham, if Google is worth $300 a share, capitalism is broken.
As of last week, the Market Climate for stocks remained characterized by unusually unfavorable valuations and modestly unfavorable market action. We continue to observe dull volume here, while bullish sentiment (a contrary indicator) has jumped in recent weeks. Few new developments last week, however – overall, we've still got market action suggestive of distribution, but enough resilience in market breadth to allow for the possibility that investors will recruit a speculative mood. For now, however, the overall character of market conditions remains unfavorable. That's not an indication of probable market direction in this specific instance, but it does indicate that on average, the market's average return/risk profile hasn't been very satisfactory in conditions that have resembled what we see today.
In bonds, the Market Climate remained characterized by unfavorable valuations and fairly neutral market action. It's notable that when the 10-year Treasury yield dropped below 4% last week, a number of analysts asserted that stocks were deeply undervalued on the basis of the Fed Model. As I've noted before, the Fed Model, though worse-than-useless for stocks, is actually a fairly good bond market model. Specifically, the Fed Model basically compares the prospective earnings yield on the S&P 500 to the yield on the 10-year Treasury. Historically, when the Fed Model has indicated deep stock market undervaluation in a period when stock yields are already low, it means, by definition, that 10-year Treasury yields are even lower. Though such signals have no correlation with subsequent stock market movements, they've tended to be good sell signals for 10-year Treasuries, indicating that long-term bond yields are probably too low to be sustained.
As usual, we don't base our investment positions on forecasts or predictions. At present yield levels, we don't observe enough yield or potential for capital appreciation to warrant a substantial exposure to long-term bonds. It's certainly possible that the economy could deteriorate sooner rather than later. But even in that case, I would expect continued inflation pressures (i.e. stagflation) rather than a substantial further easing in long-term yields. An abrupt widening in credit spreads would change my mind about that, since it would indicate the probability of credit defaults, which is normally associated with declining monetary velocity and deflationary pressures. For now, however, there's not enough evidence of deflationary pressures to warrant much exposure to long-term bonds at their present, low yields.
New from Bill Hester: Earnings Growth and the Yield Curve
|Home | The Funds | Open an Account | Account Access | Research & Insight | Site Map|
For more information about investing in the Hussman Funds, please call us at
513-587-3440 outside the United States
Site and site contents © copyright 2003 Hussman Funds. Brief quotations including attribution and a direct link to this site (www.hussmanfunds.com) are authorized. All other rights reserved and actively enforced. Extensive or unattributed reproduction of text or research findings are violations of copyright law.
Site design by 1WebsiteDesigners.