Different stocks for different reasons, but I tend to look at things on a somewhat macro basis. Recent "trends" in the past 2 yrs that I've used as a guide:
- high cash levels, low valuations, and brisk dividend growth in big tech (bought MSFT and AAPL)
- oil went down in a major way, so bought 1 stock in various areas of energy around every few months (CVX, XOM, COP, OKE, KMI)
- financials are still a relatively good value, and litigation seems to finally be slowing somewhat overall, expectations aren't all that high, yields could grow once they start passing stress tests (AIG, WFC, AXP, TRV)
- high flying growth names had corrections, and I felt that with reasonable growth and even projecting a moderate P/E in 10 yrs, they could offer a good return (SBUX, DIS)
- cyclical and limited history of good performance in automotive, but very cheap and seemingly working hard to restructure and lower costs and invest wisely (F and GM)
I work in finance, so I could come up with fun and complex models to value stocks, but at the end of the day you can make the model show what you want anyway. For me, I try to take a macro view on both industries/sectors as well as my own portfolio. For example, even though energy continues to be a good value, I don't just keep adding forever since everything carries risk. I try and hit some very loose industry weights, and keep a list of stocks/companies I like and try to pick one every time I have enough cash for a purchase (tends to be <10 purchases per yr).
Sometimes I catch something at the wrong time, and other times I get a better value (ex. energy stocks over the past yr), but overall, good to buy the reasonable value and growth as it comes available and I have room to take the risk. I tend not to sell often (I think 2 sales in the past yr), but am certainly open to it if my reasoning changes.