Ask Colin

Finding interesting companies using the AFR Rolling Year Records is OK, but how do I know if today's list is better than any other day's list?

I think that you might be misunderstanding how I am suggesting that the Rolling Year Record table be used.

In my June 2003 Shares magazine article Finding Hidden Gems, I said:

One useful way to find up trending stocks is to use charting software filters to assemble a list of stocks that are making new highs for a significant period. One of my personal favourites is to look for stocks that have recently made a new high for the last 52 weeks. There is nothing sacred about 52 weeks, or one year, but it seems to work quite well. This list is also published daily in the Australian Financial Review (AFR) in the Rolling Year Records table. Other common periods used are 26 weeks, 30 weeks and 40 weeks. These will sometimes pick up trends earlier than the 52-week filter.

What we are trying to do is find uptrending stocks. One way to do that is to scan for stocks making new highs for a substantial period. They will either be trending up over that time, or breaking upward out of an accumulation or consolidation zone and therefore have potential to trend up. This latter group are the stocks that might fit my value or growth models as explained in Shares Charting Guide No 1.

If we have charting software that allows this type of scan, we can run it every day to get the list of stocks making 52 (or other number) week new highs. In practice, I run the scan once a week and look for stocks making new 52-week new highs any time in the last 20 days.

If we do not have software with this capability, it is also easy to do the same thing using the Rolling Year Records tables in the AFR each day. The way I used to tackle this is to keep a spreadsheet listing of the stocks that came up on the 52-week new highs list. As each new table was published, I added any stocks to the spreadsheet that had not appeared before. The spreadsheet makes it easy to track them, because you can sort the list as a stock is added, so you can spot new stocks quickly and easily. However, it can all be done manually. This means that you do not have to research every stock every day, but only research those stocks that appear for the first time, if their charts are interesting.

As for whether a stock that appears today is better than one that appears next week, there is no way of knowing. You have to do your analysis of their charts and then their fundamentals and make a judgement. Even then, you could be wrong, or the situation could change. That takes us into a whole new area - how you manage a portfolio of stocks. This is a big subject, but the key ideas in my method are:

1. Take a small position in promising stocks.

2. If they continue trending, gradually build your position.

3. If they do not continue trending, sell out and move on.

4. When the trend is finished, sell and move on.