Ask Colin

(Concerning the testing of the "Dogs of the Dow Theory" and its variants), do you think the published value of Div % is reasonably accurate for this analysis or should I be looking for more accurate data? I could obtain this back history from the CRIF lab at NSW Uni, at significant cost, but if the newspaper data is close enough then I can continue on and save the money. I note that most people buying stocks on yield and price would do so straight off this list anyway.

Surely there are three issues here:

1. Selecting the stocks. The stock selection methods all seem to use historical dividend yields. These are what are printed in the newspaper. So, I see no problem selecting the stocks you "buy" in each variant method based on the newspaper dividend yield. However, one issue you have to face is dividend imputation. Should you use the grossed up yield or the raw yield? This is not an issue in the US work, as they do not have this problem. It may not make any difference in the stock selection process. This will be easy to check, because the Newspapers publishes both yields I think.

2. Evaluating the results. Historical yields are no good, because the company may increase or decrease the dividend while you are "holding" it. Also end of year yields are no good, because you have to take only the dividends that flowed to you while you "held" the stock. Some of the dividends represented in the end year yield calculation may have been or will be paid outside the time you were holding the stock. So, you are going to have to research what dividends were actually paid each year. The best source may be the Stock Exchange Journal, but your broker or the company may be able to help too.

3. You need to consider taxation - imputation credits on dividends (note the franking rate has changed several times) and capital gains tax. Assuming you hold the shares for a year and a day, you will fall into the capital gains tax system (which has also gone through a couple of changes over the years and did not exist at all before 1985). If you only hold for the year exactly, then you pay full tax on the gains. In both cases you can offset capital losses against capital gains (with luck you will not have to carry any of them forward).

One of the things you must do is not only look at whether the variants beat the "dogs", but whether buy and hold would have been better that all of them. For buy and hold, you use the All Ords Accumulation Index. The key point is that the index is pre-tax. However, if you are going to rejig the portfolio each year, you must reduce it for tax to make a valid comparison to the index, even if you ignore tax on dividends for the comparison.

Since you are taking the past and looking at how it would have gone, with a view to using the methods in the future, it suggests that you should notionally tax the gains under the regime you will face in the future.

A year ago, I might have been happier in ignoring tax on dividends completely. However, consider the new measure to refund excess credits. If you are a taxpayer on less that the company rate of tax, you will have excess credits, which the ATO will refund to you. This boosts the return on the investments.

I think you need to think these issues through very carefully.

If could add another comment. What you are taking on is a huge job. If you love the process of finding which is best, then go for it. But remember life is short. Perfection in investment is never achievable. It might be better to make the best decisions you can and get on with life, rather than spend countless hours trying to work out which one gave a fraction of a percent better return in 1982, when the future will never be like the past. It is very difficult to beat buying and holding a diversified portfolio of stocks. It can be done, but at the cost of much time and effort and at the risk of failure at worst and greater volatility of returns at the best. You need to decide whether any endeavour is worth the effort and what the payoff is for you as a person who is only here for a short while.