Ask Colin

I have just finished reading your Research Report No 1 on combing FA and TA, which I found very informative. I find it difficult to understand how one can rely on dividend yields and PE's, which are based on earnings calculated up to 6 months or more previous for the filters. The dividend yield is obtained by dividing the DPS by the present market price of the share. Hence, if the share has dropped in price, the yield would be increased and in a similar manner the PE would be decreased. This would give a false positive aspect to the stock. The situation would soon be obvious when the chart is observed, so does this indicate that only shares with a positive trend should be considered in the filter list?

I have just finished reading your Research Report No 1 on combing FA and TA, which I found very informative. I find it difficult to understand how one can rely on dividend yields and PE's, which are based on earnings calculated up to 6 months or more previous for the filters. The dividend yield is obtained by dividing the DPS by the present market price of the share. Hence, if the share has \d\r\o\pped in price, the yield would be increased and in a similar manner the PE would be decreased. This would give a false positive aspect to the stock. The situation would soon be obvious when the chart is observed, so does this indicate that only shares with a positive trend should be considered in the filter list?

There are a several issues here.

The first issue is the use of historical (past) dividend yields and PR ratios - that is ratios that use current prices, but dividends and earnings that may relate to periods six months or more ago. The same problem applies to the price to net tangible assets scan also. I use historical ratios because they are conveniently available to scan many hundreds of stocks. Prospective (estimated current or future) ratios would be better. Putting aside the difficulty that the estimates may be faulty, the problem is getting hold of them. There is no easy and economical source of the prospective ratios for many hundreds of stocks.

So, I use the historical ratios. The logic to it is that I only use them for industrial companies that provide an ongoing product or service. The assumption is that this type of company has some base of maintainable earnings. Accordingly, past ratios may be a good guide.

The second issue is that the scan is just a way of finding interesting stocks. It is the start of a search rather than the end of it. Once I find interesting stocks, there is still a lot of work to do in investigating the company to make sure that the historical ratios make sense for that company. Many will be discarded at this point, because it is clear that they are cyclical companies or have had an unusual peak in earnings and next year is likely to be different. Usually, the chart is going to tip us off about this, but not always.

The third issue is that the price may have fallen and so made the ratios look cheap. This is the whole point of the exercise. We are looking for companies that are cheap. However, while the few good stocks we are looking to find are undervalued by the market, a great number will be simply companies that the market is not expecting will perform as well this year and that is why they are cheap on the historical ratios. In the report, I took some pains to stress this and used the example of Carlovers Carwash to show what could happen.

This brings us to the third issue, which is that the fundamental scan is only a means to find a list of cheap stocks based on historical ratios. The real work is done in the next step, which is to look at the charts of the stocks that come out of the scan. You are quite right that this is very important. I spent many pages of the report taking you through how I analyse the charts looking for the few good ones. This is the most important part of the approach. Yes, I am looking for uptrends, but also to flag where breakouts would be above significant trading ranges.



Keywords: