What I use to pick Compound Income shares.

This is the second of three articles where I explain the why, what and how of my Compound Income investing. In this part I will look at the other factors I consider in addition to yield which was covered in why.

Value

Even though yield covers this to to a certain extent as often higher yielding shares tend to offer value as they have got onto a high yield by being beaten up or neglected. I do however like to check other vauation metrics when considering yield shares, to ensure that I am buying value and not just an expensive high yielder.

No discussion of value would be complete without a mention of Benjamin Graham and his focus on buying cheap Assets or Net Nets as he called them. Other studies of returns to value have also found price to book value to be a good predictor. However, I have always struggled to find many Companies meeting Graham's deep value criteria and while I wouldn't put people off using price to book, I consider it but I prefer to focus on profit, earnings and dividend based measures of value. Nevertheless in certain circumstances I use price to book such as for asset plays like property and investment trusts and for certain financials like banks.

So I consider the basic P/E as buying low P/E stocks is also generally a good idea as shown in the What works research I highlighted last time. The main issue with P/E's is that it does not take the capital structure into account. So sometimes stocks are on cheap P/E's because they are highly geared (debt) which can be both good and bad. I therefore prefer Enterprise value (EV) measures that take the full capital structure including debt into account. While there is probably no one best value measure it seems from research that EV/EBITDA comes close, so I tend to pay most attention to gross profit to asset measures and earnings yield type calculations where EBIT or EBITDA is compared to EV. I then factor these in alongside yield in my screening. However, it can also be argued that a composite value rank composed of several value factors can do an equally good job. See the research link above and the book Quantitative Value for more details on comparisons of value measures and Stockopedia which has a Composite Value Rank which is probably an easier alternative for most people.

Cover, growth, operating ratios and financial security (or Quality)

Next I want to cover quality and what I mean by it in relation to compound income shares. This comes in three forms, firstly I am interested in the quality of the income and this can be indicated by how well the dividend is covered by earnings and free cash flow. The rationale for this is from this paper by CSFB which showed that cover helped to select better performing shares in the yield space and this seems quite logical and is quite commonly used. The level of cover will also be an indicator of the security of the dividend i.e. scope for it to be maintained or increased, which is what I am looking for in the Companies I invest in. I also want to avoid those stocks that are vulnerable to cutting their dividends, which is pretty obvious, but why this  is a good idea is set out in this research paper on dividend omissions.

Another aspect of dividend quality that I am interested in, given that I am looking for Companies that can grow their dividends, is the dividend history or streak, together with forecast dividend growth. This can help me to identify trends and Companies that seem able to offer what I am looking for given the signalling inherent in a dividend policy. Some research has been published on this in the UK and while not conclusive it did show that it worked well in equally weighted portfolios and therefore is useful with smaller companies. The other significant feature of this research was how badly zero yielders as a group have performed in this extract from the research:

It is observed that a considerable portion of the excess returns available to equally weighted consistent dividend portfolios can be attributed to the avoidance, by definition, of any non-paying stocks during the period studied. Whilst there have been periods of spectacular outperformance by zero-dividend firms, over the full period of 1986-2006 they have demonstrated relatively poor returns coupled with markedly higher volatility than dividend paying firms.

Thus as a rule I tend to exclude zero yielders from my investment universe and also tend to use a low yield threshold to act as a sell signal / discipline.

The second aspect of quality I am looking for is the quality of the business which is generating the cash flows to pay the dividends. In this area I have tended to look at Return on Capital Employed and operating profit margins. More lately though, since they developed it,  I have also been using Stockopedia's Quality ranking which is a composite measure for this factor, plus a QV rank where it is combined with their Value composite. Research which supports the use of quality screens generally and in income portfolios has been produced by GMO and Soc. Gen. respectively and in relation to value in this academic paper by Robert Novy-Marx.

The last factor related to quality that I address is that of finacial security of the Company or essentially the state of the balance sheet and how it is moving. This is also covered in the Quality research from GMO & Soc. Gen. In addition this recent paper from Cliff Asness et al and another one that explored Buffet's performance also highlight its importance and the value of avoiding junk. I also particularly liked the paper also by Asness et al about betting against beta. 

Therefore, I prefer lower gearing (debt to equity) and further measures that I use to assess financail security / risk include interest cover, the Altman z score and the Piotroski score plus the actual levels of debt to equity. Again this is designed to weed out the financially weaker companies that might be more vulnerable to not growing or cutting their dividends if things don't turn out as expected. Since I am looking for yield and value it would be acceptable to stop there and select quality, value and yield candidates from the above and indeed I do run screens with just those inputs in the score.

Momentum

However, there is evidence that momentum in various forms such as price movements and estimate revisions can also be predictive of future performance. In addition there is research which shows the benefit of combining value and momentum factors to help smooth out some of the fluctuations in performance between the two styles.

Consequently I also look at rankings including estimate revisions, which if positive, makes it more likely that the forecast growth and trend might be achieved and continued. Or alternatively that a good value / yield  stock with poor price performance might be worth a look if the estimates have started to rise and the market still hates it.

While I'm less keen on price momentum it can also help to avoid losers or value traps, especially when combined with looking at earning revisions which if also poor tends to keep me out. Avoiding losers is often just as important as picking winners. The converse is where stocks have performed well, then obviously the value and yield will tend to be less attractive and you then have to watch the momentum factors carefully for signs of trouble. Once again Stockopedia has done much of the work for you here with their combined momentum rank, although personally I prefer to split the momentum factors as discussed above.

In addition, I also screen for three year underperformers on the basis of mean reversion (see What has worked paper from Why Compound Income?) to see if I can find any attractive yield stocks that may have recovery potential or seem to be turning around as evidenced by earnings upgrades and or price performance picking up.

Despite my reservations I have found it ueful to pay attention to these factors as often it can be an early warning sign. In addition you can get a new stock coming up on yield screens only to discover that it has just collapsed on the back of a profits warning. If this is the first one out of the blue I find it is then generally worth ignoring that stock for up to a year as often more profit warnings or downgrades follow until the Company and analysts adjust to the new reality.

Summary and Conclusion

My approach would probably be categorised as income growth investing. I describe my investment strategy as a Quality Value and Yield based strategy focussing on valuation, yield, dividend cover and growth, business characteristics and financial security. In addition I pay some attention to momentum factors, especially earning revisions and price momentum and utilize quantitative techniques (including Stockopedia) to help identify new ideas and monitor my portfolio.

The bottom line is that I am aiming to grow my capital and income in real terms over time. I therefore look to Compound Quality Growing Income or CQGI which is the essence of Compound Income?.

PS. If you find all the research papers a bit heavy going I have just added a post about a Simpler explanation of risk factors here.