From Investment Gurus by Peter Tanous:

Peter Lynch : The problem with technical analysis is that somebody could love the stock at 10 and hate it at 6. But I have traditionally liked one formation. The stock goes from 50 to 8, then is goes sideways for a few years between 8 and 11. Now if something goes right with this company, the stock is going north. These make a nice research list. You look at stocks that have bottomed out. Its like trying to catch a falling knife, you want the knife to stick in the wood. When it stops vibrating, you can pick it up.

Laura Sloate : In 1973, I couldn't find any stocks that were cheap enough. Management was on my case since I wasn't generating any commissions. At that point, I realized that if you worked for a big company, you would be controlled by the management.

If the average return on invested capital is less than the cost of the capital, you know its a poorly managed company. If we look at our stocks through 91-94, 12 of the 15 stocks that we sold because of a sell discipline of selling at 15% decline, 12 of these were higher six months later when we originally bought them, and a couple were real winners. So, we modified our sell discipline, we don't let it go down and not do anything. Its an evolution of process based on examination of results. Statistically we found that in 80% of the cases, if we had kept the stocks we sold six more months, we would have made money. Most of the bad stuff is in the price; we sometimes bought prematurely.