Chris,

I have evolved a process of generating watchlists by analyzing the parameters that positively correlate with one year price increases in the top 100 price gainers of each year. IBD publishes the top 100 price gaining stocks of each year in the first edition of IBD each January. I found four factors that correlate positively with price gain: Liquidity, Demand/Supply, Margins, IBD Composite rating. I created a process to measure these parameters and rank stocks based on them. I also wanted a repeatable process. Subjectivety is unavoidable but you will see some science in my process below.
My process starts with downloading all stocks trading above $5 in the marketsmith database. This is usually more than 4,000 stocks. I create three local ratings for each stock. First is a liquidity rating (1-99). Liquidity is defined as 50-day average volume times close price. A rating of 95 for example says the stock is more liquid than 95% of all stocks in the database. The second rating is for demand supply ratio. Demand /supply is defined as 50-day avg volume divided by the float. A 1-99 percentile Demand/Supply rating is then calculated for each stock. I then sum pretax margin and Return on Equity and compute a 1-99 margin rating for each stock. Finally I use the IBD 1-99 Composite rating. These four ratings are then summed into a Combination rating. Stocks with a high Combo rating are of interest (top of a ranked list).

The reason high Combo ranked stocks are interesting is the following. Highly liquid stocks have to be bought primarily by institutions (the I factor in CANSLIM). High demand /supply (the S factor in CANSLIM) means the demand is high compared to the number of shares available for trading. High margin rating is an indication of profitability and effectiveness of management. High IBD Composite ratings is an indication of all CANSLIM traits.

This process isn't as difficult as it sounds as most of the process is accomplished with an excel macro. I have tried correlating other parameters with one year price gain with no clear positive correlations.

Now I filter out all stocks that fail to meet minimum criteria: Price above $15, Liquidity above $10M/day, Demand/Supply rating above 45%, Price above 200-day, Combo rating above 300, Accumulation/Distribution rating C or higher and the average of the last two quarters EPS increases 25% or more. The stocks that make it through this process numbers around 75 at which point I shift to a manual chart and fundamentals inspection. I look for acceptable CANSLIM traits, a rising chart, a buyable base where price is not extended, I check to see if the stock is owned by a quality mutual fund. Finally a short list makes it to the watchlist complete with buypoint. Every base has to meet minimum length and maximum depth requirements. For example a cup and handle must be at least 7 weeks long correcting less than 30% with the handle no deeper than 15% and situated in the upper half of the base. There also must be more accumulation and support bars in the base than distribution and stall bars. Quality mutual fund ownership is measured by determining whether one or more of the 20 mutual funds in the IBD mutual fund index owns the stock.

When you have inspected millions of charts vetting a single symbol can be done very quickly much less than a minute per symbol. All of this work is done on a weekend.

Short candidates are found more haphazardly. To be an acceptable short candidate it needs to have been a prior leading stock. As such it would have shown up in the above process at some point in time. The reason for this is if you want the stock to drop a lot it needs to have risen a lot. So when a leading CANSLIM stock falls below the 50-day it is inspected to see if it might be shortable some day. I don't short thin stocks so the first requirement is that they trade above 1M shares per day. The watchlist process then comes down to reviewing the collection every weekend for a possible technical set up with a proper short point.

I am not sure what you mean by price change > 0. I am guessing the volume alert criteria. The volume trigger alert watchlist includes every stock on the regular long watchlist but the technical screening parameters are loosened to include stocks not at a classical buypoint but in the range where a viable volume alert buypoint could materialize. A valid volume alert trigger needs to meet the following criteria: price closes up on the day (price change >0) with volume greater than any down volume during the prior 10 days. Price needs to ideally be at or really close to the 50 day or 10 day moving average.

When you think about it there is a lot going on behind the scenes vetting stocks to go on the weekly watch lists. These watch lists are the ones that I use for my personal trading.