Discussion-board regular Terry Harper outlines his high-yield investing style.
I have made many contributions to the two High Yield discussion boards (here and here) over the years, in the course of which I've set out my method of running my own High Yield Portfolio (HYP).
It has been suggested that it would be a good idea to collect all my various principles into a single article, so that reference to them would be made much easier and convenient.
My portfolio began in essence when Personal Equity Plans (PEPs) were introduced in 1987 and has essentially followed a traditional criteria for income share selection -- that is, choosing a mix of shares with higher yields from diverse sectors.
By 1997, I realised that two holdings represented more than 10% of my portfolio and I realised that, while I could do as I liked, a unit fund would not be allowed to get into this situation.
Consequently I established my first personal rule: that 10% should be the maximum weight of any one holding and any share exceeding that limit should be trimmed back by about 20-25%, that being an economical trade for me at that time.
Subsequently, as my portfolio grew in numbers and in value, the 10% limit was changed to twice the median holding value (when I had more than 20 shares) and then to 1.5 times the median value when the number of shares exceeded 30. Just what anyone else should use depends on their own circumstances, and the size of their portfolio. This post of mine gives some of the background.
By 2000, the frequency of my trimming began to increase as the markets became more volatile, and so I had to decide where to invest the proceeds of my trimming and accumulated dividends. Up to this time, I had chosen either the share with the highest yield or the one with the lowest weighting.
I came up with a method that combined the rankings of yield and weighting to arrive at an order in which shares should be topped up. The method is explained in this post and expanded upon here at a later date.
As things developed, it became obvious that some shares had outgrown their usefulness in terms of income generation, because their yields were now well below that of the market. Accordingly, I developed my own criteria for disposing of share holdings completely. The criteria are:
1. The share has grown to such an extent that its yield is now less than about half that of the market, typically less than 2%.
2. The company has stopped paying dividends and is unlikely to resume paying them in the foreseeable future.
3. The company is demerging and spinning off shares that would not be selected for an HYP. Such shares might also be less than an economical holding value.
Companies sometimes reduce (or 'rebase') their dividends. If the dividend yield is still in the acceptable range above that of the market average, then I would keep and possibly add to the holding. An example is RSA Insurance (LSE: RSA), which disposed of its life business and accordingly reduced its dividends in 2002. I still hold RSA. However, I would not normally buy shares in a company that's not currently paying a dividend.
One of the problems in tracking the performance of a portfolio is the effects of adding capital, reinvesting dividends and withdrawing cash. After a number of unsuccessful attempts, I finally decided that the best method would be to unitise my portfolio. Here is a useful thread on the subject, where this post by me gives a worked example.
There are two ways of dealing with dividends. You can either let them accumulate inside the units, just like the accumulation units of an OEIC (open-ended investment company), or you can buy extra units with them, as per reinvesting in the income units of an OEIC. Here is a post that explains this concept further.
If you wish to compare your portfolio’s performance against an index such as the FTSE 100 (UKX), then using income units makes them directly comparable. If you follow the accumulation unit route, then you would have to use the Total Return version of the index.
A further advantage of using income units is the ability to calculate dividends from the portfolio as dividends per unit. These can then be compared directly with the Retail Price Index (RPI), to ensure that you are receiving an income that is growing at least as fast as retail prices. An example of this calculation can be seen in this post, where I show dividends per unit and RPI from a common base.
I hope that, by collecting these various links together and providing some historical background, fellow Fools can borrow my ideas freely in their own HYPs.
> Terry Harper contributes to the Fool's discussion boards under the Author name tjh290633. He owns shares in RSA Insurance.
