The problem is how to separate the companies that will flame out after 5 years of hyper growth from those that will profitably reinvest their retained earnings into greater and greater dominance. That will always be the hardest part of investing.

This also speaks to why PE ratio in isolation is a useless valuation metric. Dying companies often have very attractive PE ratios, while companies that will produce 30+ years of compounding cash flows have very high PEs.

This realization has given me more sympathy for people who pay crazy prices for hot stocks. Sometimes the mania is actually justified—if they hold for 20 years and ride the compounding growth.

But truly outstanding companies ("compounders”, in investor-speak), generate cash for decades longer than forecasted. The difference between forecast and reality can be *massive*.

The reason for it comes down to the mechanics of the DCF models that analysts use to make their forecasts. These models require a fixed period over which to forecast growth, typically not longer than 10 years, with any growth after that seen as negligible.