Wizard wrote:Gengulphus wrote:Wizard wrote:Unless somebody practices SI in a very purest way then when selecting purchases I think we all take a view on the shares we buy.
I think we all take a view on the shares we buy - full stop! Even practicing "SI in a very purest way", e.g. as pyad explained it in the quote idpickering has dug out, doesn't prevent one taking a view - it just prevents one basing that view on long-term forecasting. One can still take the view e.g. that "this company has consistently grown its dividend over the last 5 years and is forecast to do so again this year - that's good enough for me", or (more appropriately for this thread) "this company has shrunk its dividend over the last 3 years and is forecast to do so again this year - no need for me to waste any more time on it".
Well then I guess we could get into debates aout what is "long term", I seem to recall that the view of the economists is something along the lines of "in the long term we are all dead"
Well, it was pyad who wrote the quote about SI meaning ignoring long-term forecasting - so probably best to ask him what it means in that quote...
But in the version of SI that I use, it means further into the future than I think it's possible to forecast the fortunes of the company reasonably reliably, which depending on the company means 1-5 years, with most companies being 1-2 years and very few getting up to 5.
Note that that's
not the fortunes of the company's business, but of the company. In particular, it's often the case that a business can clearly be seen to be in long-term decline: what isn't anything like as foreseeable is how successfully companies in that business will move on from it. E.g. there's no real doubt in my mind that shopping in physical supermarkets is slowly becoming less important and will continue to do so, and all of the supermarket companies are producing internet shopping alternatives in various different ways. But which of them is going to do best at that and find the best uses for the physical supermarket properties as they are slowly released from the business are crucial questions when looking at the long-term futures of the companies, and I don't think that is very forecastable at all.
Wizard wrote:What puzzles me a little then is what IIRC was a concept used by Luniversal, that of a "danger zone", where a yield was too high to be believed to be sustainable. ...
Well, it doesn't puzzle me - but that's only because I quite simply don't believe in the concept at all!
What I do believe in is that a high yield can be, and quite often is, associated with an unsustainable dividend. But there are two crucial differences from Luniversal's "zones": firstly, I believe that's true of any high yield, and secondly, I don't believe it's automatic. I.e. the "zones" concept essentially says that a moderately high yield implies the share's dividend can be treated as sustainable and a very high yield implies that it should be treated as unsustainable, while I believe that both moderately high and very high yields imply that the sustainability of the share's dividend should be investigated and a decision made based on what the investigations reveal.
Suitable investigations (or "dividend safety factors") are things like dividend cover, debt levels and how they're changing, the management's record at setting dividend levels sensibly, looking for foreseeable events that are big enough to significantly affect the company's ability to pay dividends, and company size relative to the plausible size of less foreseeable events. E.g. both Shell and Unilever are megacaps and both could suffer all sorts of unforeseeable problems - but it's much more imaginable that Shell might suffer a tens-of-billions disaster (as BP did in 2010) than that Unilever might. All of those can be considered without doing long-term forecasting, and so are compatible with SI.
Wizard wrote:As an example TalkTalk has a very attractive yield (well into the "danger zone" I believe) at present because of a significant fall in share price, but I do not think there is any concrete statement or forecast of a dividend cut. How does a share like that fit? As it happens it is off my list for reasons of the level of gearing so I have not considered whether I think this is a long the share price or the market signalling a reduction in dividend.
It's off mine as well, in my case because with BT and Vodafone already in my HYP, I'm not looking for any more telecoms companies. But if it were on my list, I would basically be doing investigations as described above to try to see how easily the company is currently / in the near future paying its dividends, not thinking in terms of the market 'signalling' things.
Or at least, not in terms of it 'signalling' useful things. The yield being high does mean that the price is low relative to what the company is paying out to its shareholders, and the price being low does indicate that something about the opinions of market participants who are actively considering buying or selling the share. And some of those market participants base their opinions on what the company is paying out to its shareholders - but a lot are basing them on other things entirely, such as trying to analyse the behaviour of other market participants to work out where they're taking the share price in the short term. That leads to positive feedback loops like share prices rise => those shareholders see other market participants as mainly buying => those shareholders think the price is likely to rise => those shareholders buy => share prices rise because of the extra buying pressure, and the corresponding one with "rise" and "buy" replaced throughout by "fall" and "sell". And those feedback loops can lead to prices becoming irrationally high or low - with the tech boom at the turn of the century being an extreme example of the irrationally-high case, and the example tjh290633 gives from time to time of Stagecoach in late 2002 being a pretty clear example of the irrationally-low case IMHO.
So I could think of a high yield 'signalling' that one of a number of things is happening, with one of them being that the dividend really is likely to be cut, another being that the share price is being driven irrationally low by such feedback, and there may well be others I haven't thought of for this post. But with the first indicating that it's a poor HYP purchase and the second that it's an excellent one, that 'signal' isn't very useful! I need instead to look for things that should help me to distinguish between them, such as the sorts of "dividend safety factors" that I mention above to see if they indicate that the dividend is unlikely to be cut.
Or something that indicates that the share price is being driven irrationally low might be used, such as the contrarian indicator of observing large numbers of people saying they think retailers are an uninvestable sector... But I have to admit I mostly don't dare act on such indicators, and very possibly am missing out as a result. For instance, I was getting that impression about resources stocks around a year ago, but basically didn't act on it - and in retrospect it would have been an absolutely brilliant time to have loaded up on them! Though that may well be drifting off-topic, being more of a capital-oriented verdict than an income-oriented one, and so I'll leave it at that...
Gengulphus