HOW MY STOCK PICKING MODEL WORKS
I developed a stock-picking model when I was in graduate school to take advantage of the bullishness the market was exhibiting at the time. The premise was to invest in high beta stocks while trying to limit my downside exposure in the event of a stall or downturn. By using the quantitative steps in the model, stocks are selected that are experiencing sustainable price momentum. The model is a multistep screening and high-grade process that goes something like this:
First, create a universe of about 200 stocks that have demonstrated strong price appreciation and earnings growth in the last 12 months. You can find some pretty good free and easy to use screening tools on the Internet such as Quicken, Stockpoint or MSN. I use the following screening criteria to build my universe of stocks:
Screen #1: Stocks making new highs with IBD EPS rating of 90 or higher
Screen #2: Stocks within 2% of their 52 week high with a 52 week price appreciation > 300%
Screen #3: Stocks within 5% of their 52 week high with a 52 week price appreciation > 150% and EPS growth of 25%
Screen #4: Stocks within 5% of their 52 week high with a 52 week price appreciation > 125% and P/E < 50
Screen #5: Stocks within 5% of their 52 week high with a 52 week price appreciation > 50% and P/E < 15
Screen #6: Stocks within 10% of their 52 week high with a 52 week price appreciation >100% and P/E < 100 and 3 year earnings growth of 50% and 5 year sales growth of 25%
Screen #7: Stocks with Investors Business Daily Ratings greater than 95 EPS and 95 RS.
Screen #8: Stocks with Investors Business Daily Ratings greater than 90 RS and PE less than 20.
Each screen will yield about 20 â 25 stocks. Rank these stocks on the following criteria: Price Appreciation, Price Appreciation divided by trailing 12 month P/E, Price Appreciation divided by forward 12 month P/E. Weight each criterion equally and rerank the database. This process favors stocks with more reasonable valuations and weeds out those with no earnings.
Take the top 20 stocks and rank by 12-month revenue growth.
Take the top 10 stocks and run a time series regression analysis on the daily prices for the last 12 months and rank by the highest r-squared correlation coefficient.
Take the top three stocks in order and perform due diligence to determine if there were any one time non â operating factors that affected the data just analyzed (asset sales, lawsuits, financing, etc.) or if there is any pending news of significance that could upset the applecart. Select the highest ranked stock that clears this hurdle.
Buy this stock. In a typical bull market, the stock will, on average, achieve a 15% gain within 4 to 6 weeks. Sell the stock and repeat the process. Why sell so soon? Well there are ever changing phenomena going on in the market that could make your selection criteria quite different a month after the signals told you to buy this stock. The theory here is that you are selling a potentially "tired" stock and trading it for a "fresh" one.
What this process is trying to do is to select a hot growth stock that has a little more juice left in it to get you that last 15% without being so hideously overvalued that it could drop like a rock. I donât think I need to buy stocks with extended valuations to make a quick profit. There are stocks out there with good momentum that arenât bad to hold if I make a wrong decision. I think my model finds them. My model has been successful in protecting me from real lemons. Preservation of capital is always important. Buying companies with real earnings protects me in the down markets. We all work hard for our money. It makes no sense to give it away. Thatâs why I believe itâs important to buy stock in companies with real earnings.
