Our 'Dogs of the FT30' mini-portfolio, based on the principles expounded by US fund manager Michael O'Higgins, has come up trumps this past year. That makes four years out of the six we have been keeping records that it has outperformed the broader market (see table).
The saving grace so far this calendar year has been the takeover of Logica, effective from August, which resulted in a 70 per cent gain in this particular dog and more than offset another poor performance from Man Group.
Overall, the five-stock portfolio is up 14.1 per cent since the turn of the year. Added to the prospective yield at the outset, this gives a total return of around 22 per cent, versus a 7.5 per cent gain on the same basis for the FTSE 100 index. The mechanically generated 'nap' selection of RSA Insurance produced a 5.7 per cent gain which, together with its projected yield at the outset of 8.7 per cent, suggests a total return of 14.4 per cent.
After Logica, the next best performer was Ladbrokes. It scored a gain of 37.7 per cent over the period. At the time of writing, Logica was about to be delisted. As a consequence, it will be replaced in the FT30 index with a new constituent, normally the largest non-constituent by market capitalisation.
How the Dogs of the FT30 have performed (2007-2012)
| Year | Initial yield (%) | Capital performance of (%): selections | Capital performance of (%): FTSE 100 | Capital performance of (%): single stock | (company) |
|---|---|---|---|---|---|
| 2007 | 4.8 | 6.8 | 1.7 | 15 | (BT) |
| 2008 | 4.1 | -50 | -34 | -5 | (RSA) |
| 2009 | 11.1 | 41 | 22 | 72 | (Logica) |
| 2010 | 7.2 | 6.7 | 10.5 | 5 | (Ladbroke) |
| 2011 | 5.5 | -7.5 | -1.6 | -18.2 | (RSA) |
| 2012 | 7.9 | 14.1 | 4 | 5.7 | (RSA) |
How the dogs approach works
In brief, the background to the system is that in each of the last six years, normally at the end of October, we have selected a portfolio of stocks for the coming year based on the O'Higgins' 'Dogs of the Dow' method, using the FT30 share index as a substitute for the Dow Jones.
The original dogs approach was a yield-based passive investing system for the Dow Jones Industrial Average. The system was outlined in Mr O'Higgins's book Beating the Dow (Harper, 1992). It has generated outstanding results, particularly when investment income is taken into account.
Investors Chronicle's ‘Dogs of the FT30' system works in a similar way. One reason for looking at a system for the UK using the FT30 - an index that is relatively little used today - is the similarity between the FT30 and Dow 30. Both are indices of big blue chips, change constituents infrequently, and are averages of share prices rather than capitalisations.
The system takes the 10 highest-yielding stocks in the Dow Jones index (or, in this case, the FT30), and then invests an equal amount in the five that, at the time, have the lowest absolute share prices in dollars (or, in this case, pence). The selections are then left undisturbed for 12 months.
After a year has elapsed, the next job is to calculate the capital gains or losses, tot up the dividend income banked in the meantime, perform the exercise again, and change the portfolio where necessary to reflect the new year's system selections and the need to reinvest the accumulated dividend income. Then sit back for another year. And that's more or less it. If any of the five selected stocks are taken over in the course of the year, the performance used is taken from the price at which they exited and the cash released held until the anniversary review.
How to select a single dog
If you only want to pick one stock to invest in using this system, there is a method for doing that too. Simply perform the selection as described above and then pick out the share with the second-lowest absolute share price. O'Higgins described this as the PPP stock (penultimate profit prospect). In our system I am calling it the SSS (single share selection).
What seems to be emerging from the results of applying this system to the FT30 is that the portfolio thus selected tends to hold its own in term of capital growth alone, but beats the index substantially (perhaps by an average five percentage points a year) if dividend income is taken into account. It also seems to work better at some points in the market cycle than others. For example, it performs particularly well in the recovery phase after a bear market trough, but rampant speculation-driven bull markets are bad news for the system.
We now have data based on the FT30 version of the O'Higgins method going back to 2004. Beginners' luck meant that the best year for the FT30 version of the system in both absolute and relative terms was 2004-05, when the stocks selected at the beginning of the period gained 117 per cent, with an 11 per cent average yield on top, compared with a 24 per cent rise in the market.
In 2005-06 the dogs gained 5 per cent while the wider market was up 11 per cent, but 2006-07 again saw good performance with a rise for the dogs of 42 per cent versus the market rise of 16 per cent.
The table shows in more detail how the system has done since then. The flipside of the good performance in some years is that years of bad bear markets are just as bad for the system as raging bulls. The performance in 2007-08 was dreadful, but it set up the conditions for a good showing in the following year. Similarly a poor showing in 2011 was offset by some impressive statistics in 2012.
What drives the system seems to be that it picks the stocks that the market has either written off, forgotten about, or that it believes may be planning to cut their dividends. Because the market often overdoes the gloom, stocks like this are frequently priced too low. Sometimes, as with Logica this year, corporate buyers see more value in the shares than investors do, and the company is taken over at a handsome premium. But the system only works if it is pursued with disciplined inactivity, without any tinkering or second-guessing.
The dogs of 2013
So what are the system's choices for the coming year and why are we updating it now? To answer the second part of the question first: traditionally we have always updated the selections at the end of October. Last time round was the exception (the feature was published in early January 2012), so using the data from the end of October 2012 gets us back into the normal historical pattern.
As far as the new selections are concerned, the table shows the current constituents of the FT30, ranked by historic yield. The five stocks highlighted in bold are those that the system would select at present on the basis of their historic yield.
These are, of the 10 highest-yielding shares, the five with the lowest absolute share prices. The average projected yield of this year's selections is 8.6 per cent, mainly because of Man Group's whopping 17.8 per cent yield.
This year's selections differ only slightly from last year's, namely Ladbrokes, Man Group, Vodafone, RSA Insurance and BAE Systems. The defence contractor makes it into the list by a whisker, ahead of Tesco. RSA is again the single stock selection, being the share in the group with the second lowest share price in absolute terms.
Finally, remember our usual caveat that the system is purely mechanical and should not be pre-empted or interfered with. Fundamentals, forecasts, charts, consensus views of analysts or any of the normal stuff of investment decision-making does not come into it.
As I've observed before, it's not the best system for either ultra long-term or ultra short-term investors, but perfect for lazy, just-retired ones like me.
