The debate over whether or not the free market system is rational has been around probably since free markets themselves. Academics, unable to find time to invest because they have to think about academic stuff that's detached from reality, throw in the towel and try to prove that markets are completely irrational. Instead, they invest in well diversified funds that track the economy as a whole. On the other hand, investors such as Jim Cramer or Peter Lynch have consistently outperformed markets, disproving that market variations are unpredictable/irrational. Ironically, both may admit the markets are irrational.
The focus of this article is the rationality of investing in very "expensive" stocks; specifically I'm interested in Salesforce.com (CRM). Obviously, "expensive" is relative when the stock valuations are 5-10X book value, 80X+ P/E, etc., yet the price keeps creeping up. Salesforce has brought a solid idea into the market, yet it's hype is very reminiscent of previous bubbles. The idea has been around as long as salesmen: try to spread word about your products to as many potential buyers as possible.
Still, ask any CRM investor to explain exactly what CRM does, and it will probably take them a while to respond. On their website, you see the word "cloud" everywhere and click on the products tab and you see 15 "different" products that you can buy. Investors absolutely love fancy sounding services, just read Lynch's "One up on Wall Street". One issue that is unclear is what portion of CRM's subscribers come back and repeatedly use their services. With a hot new product, everyone will want to try something, but unless it actually increases their profit, it's doubtful they come back.
The most shocking aspect of CRM's valuation is how rationally irrational it is. In the 90's tech boom, companies fudged earnings reports to attract more investors. Not so for CRM. They explicitly tell you that they lost $0.09 per share in 2011 and project to lose $0.51 to $0.55 per share in 2012 (source: CRM's last 10-K). The losses occur due to the heavy stock-based compensation somehow excluded from the cost of ordinary operations.
Thus, if you're a CRM investor, your money is being used to pay the workers. There has been zero benefit for shareholders since 2009 when the company was last profitable. However, as this previous SA article mentioned, these opinions are all of little value as long as the herd of investors sees booming revenues. It is impossible to predict when CRM falls back to it's appropriate level. I suspect it may take a full year or two before enough companies have tried the service to realize that they're not actually generating that many new customers from it. At that point, the fall will be hard, but so it must be in irrational markets.