I prefer companies with low P/E ratios, relative to the market. That usually means they trade below 15 times forward earnings per share, making them undervalued. Since the market declines of late 2008, P/Es have dropped. The iShares Russell 3000 Index ( IWV) had an average P/E of 10.8 at the end of 2008. I am finding many promising stocks trading below 10 times 2009 earnings. In P/B ratios, I look for a slight premium, indicating investors see value in the company's assets. The Russell 3000 has an average price-to-book of 1.5. A P/B ratio below 1.0 may mean the market sees trouble ahead. Garmin ( GRMN), which has been a winner for me three times in the past year, has a solid P/B of 1.93. An important criteria for me is return on equity, which tells us how profitable the company is relative to shareholder equity. I'll use blue-chip stock IBM ( IBM) as a benchmark. Big Blue has some of the best returns going: Its ROE is 49.4% for the last 12 months. One of my recent picks -- and a previous winner for me -- is Caterpillar ( CAT), which has a solid ROE of 42.4% for the past 12 months. You won't see much return without evidence the company can generate cash flow. So those numbers usually merit a look. My frequent pick and win Halliburton ( HAL) had $973 million in cash and equivalents on its balance sheet and free cash flow of $342 million at the end of September. The fundamentals help me find a pool of potential picks that are undervalued but in a good position to rise. I turn to technical analysis for evidence that the stock is picking up momentum. I compare the 21-day moving average to the 50-day or 200-day moving averages. Archer-Daniels-Midland ( ADM) won a hefty $7,650 for me in November, just as its 21-day moving average overtook the 50-day average.