"Returns are not inherent to an asset class; they result from the fundamentals of the underlying business and [emphasis in the original] the price paid by investors for the related securities"
Seth A. Klarman - from the Preface to the Sixth Edition of Graham and Dodd's Security Analysis
I too often read statements like “stocks return an average annual return of 8%-10% over time”. Does that mean the long-term returns I should expect from equities are the same today as they were in 2009 when prices were 50% lower? If you believe that long-term fundamentals ultimately determine value, then it logically follows that returns are a function of the price paid – or else you’d be arguing that future fundamentals are somehow a function of the price you paid (side note to nit-pickers: in a very narrow set of circumstances George Soros essentially makes just such a counter-intuitive argument with his theory of “reflexivity”, but this is really centered around the dynamics of asset bubbles.)
Klarman's quote also points to one of the biggest problems faced by individuals working at institutional investment management firms – they specialize by asset class, and usually even more narrowly in subsets of a given field. It’s not that these individuals don’t understand valuations… but let’s say you are currently a 33 year old technology stock analyst making $400,000 per year living in an expensive house, with two young kids about to enter private school and a wife who stopped working when the kids were born. Have you read a lot of stories about guys like this quitting in droves because their asset class is overheated and they expect disappointing returns over the next five years?
Keep this type of context in mind any time you hear people talking more about relative valuations than business fundamentals. Of course, specialized knowledge of an asset class could help in an attractive valuation environment. But, if the market has bid prices across an asset class up to stupid levels, being smart on that field isn’t going to help much if you don't have the good sense to just avoid being long.
Think about all of the technology and biotech research analysts on Wall Street: often degreed engineers or doctors, these very high IQ individuals write exhaustive reports about the minute technical details of routers, software, new drugs, etc. They painstakingly pull the outputs of their analyses through market sizing models and competitive matrices to make detailed calculations of projected future performance. Then price targets are derived by applying simple market valuation multiples or DCF analyses, which built with inputs from other overvalued securities. Whatever errors are committed in the forecasting process are compounded (yes, in a technical sense) immensely when the punchline reads something like "we think this should trade in line with other cloud computing peers, currently at 40x estimated 2012 earnings" or, even better, "at 30x 2012 earnings, XYZ represents a compelling value at a 25% discount to its peer group average of 40x".
It's not really "garbage in, garbage out", it's more like making a prime, dry-aged filet mignon into cat food.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.