Bad Advice

Google Boggle

One analyst says it’s worth $415 a share. Another says $650. Another $601. Don’t listen to any of them.

Click here to read more of Henry Blodget’s Bad Advice.

Google reported its quarterly earnings last week, which means that the 29 brokerage analysts who follow the stock have just revised their financial projections and adjusted their price targets. As ever, Google’s newly revised value is in the eye of the beholder—the most pessimistic of the analysts now believes that Google’s stock (currently at around $470) is worth $415; the most optimistic, $650. The analyst at Bank of America, meanwhile, believes it is worth exactly $601.

Who’s right? Which target should you believe?

None of them. No analyst, no matter how talented, knows what Google—or, for that matter, any stock—is really worth.

To understand why this is so, you need to understand that price targets are not precise scientific facts but extremely subjective estimates. To calculate price targets, analysts must make numerous assumptions, each of which has a major impact on a stock’s estimated value. For a variety of reasons, Wall Street and the investment media love pinpoint price targets—they are simple, precise, easy to understand, legally defensible, and reassuring—but the truth is that the range of reasonable valuations for a stock like Google is wide enough to fly a 747 through.

To appreciate this, it helps to understand some basic valuation theory. The value of a stock, theoretically, is the “present value of future cash flows.” Loosely translated, this is what all the cash a company will pay to shareholders from now until the end of time would be worth if it were delivered in one lump sum today.

To determine the “present value of future cash flows,” an analyst needs to know the following:

1) the amount and timing of the future cash flows2) an appropriate “discount rate” with which to determine what the cash flows are worth today. (Thanks to inflation, risk, and opportunity cost, a dollar expected to be received in a year is worth less than a dollar delivered today, and a dollar expected in 10, 20, or 100 years is worth a lot less than a dollar today.)and, usually,3) A “terminal multiple” with which to value the cash flows that will be received after an explicit forecast period (usually five or 10 years).

Once the analyst knows these things, calculating present value is a matter of math. The trouble is that no one knows for sure what a company’s future cash flows will be, or even what the most appropriate discount rates or terminal multiples are. As a result, the analyst has to estimate. (To get a sense of just how much these estimates affect present-value calculations, please read this previous piece, “The Folly of ‘Cheap’ and ‘Expensive’ Stocks.”) To make matters worse, the range of reasonable estimates changes as the industry and company evolve and the prices of other securities change. For a company growing as quickly as Google in an industry evolving as rapidly as the Internet, it is not an exaggeration to say that the stock can be “worth” just about whatever the analyst wants it to be.

This is not to suggest that analysts are “making up” or “manipulating” price targets. On the contrary: In the atmosphere of fear and trembling that followed the Wall Street research scandals (starring, among others, me), analysts do their valuation work so diligently that even finance professors would be proud. Today’s analysts are so punctilious, in fact, that they are now often less useful to investors because they’re so scared of being viewed as “irresponsible” that they rarely say anything interesting. (Do you really want to read yet another report suggesting that Google is worth about $600? Neither do I. But I’d love to read an intelligent one arguing that Google is headed to, say, $50—or, for that matter, $3,000.)

Analysts develop price targets by spending hundreds of hours building intricate financial models with detailed projections of revenue, expenses, margins, taxes, share dilution, and other important metrics. The estimates are based on the past performance of the company, the industry, and the economy, but what they all have in common is that they are at best educated guesses about how the future will unfold. Because the future is unknowable, even the most careful value calculations have a large margin for error.

As financial reporting and automated spreadsheets have become more advanced, moreover, the sense that price targets and other similar estimates are, in fact, estimates has largely disappeared. Bank of America’s Brian Pitz, for example, painstakingly concludes that Google is worth precisely $601 per share—after making several subjective assumptions that, slightly tweaked, could easily have pinpointed its value at, say, $428 or $642 (or, for that matter, if Brian were in the mood to forecast extremes, $189 or $1,033). The $601 target is no doubt of great comfort to Bank of America’s legal and compliance departments, because the diligence, precision, and reasonableness of Brian’s calculations are indisputable. Unfortunately, the faux-precision implied by the target is a mirage.

 The safest way to think about stock values is to recognize that even the best analyst is the equivalent of a blindfolded hunter shooting at a moving target a hundred yards away. Given the extreme difficulty involved in hitting the target, the analyst’s gun of choice should not be a rifle or laser beam. Rather, it should be a sawed-off shotgun.

Although Wall Street compliance departments will likely continue to insist that analysts create pinpoint price targets, smart investors should never forget what they really represent: the midpoint of a wide range of subjective guesses. In Google’s case, even the Street-wide range of $415 to $650 is probably too narrow: It’s reasonable assuming nothing radical happens, but in the stock market, “radical” events happen more often than they should statistically be expected to.

As John Kenneth Galbraith noted, there are two kinds of forecasters: those who don’t know and those who know they don’t know. If you must be a forecaster, at least be the latter kind.