Page 1 of 2

Factors for future increasing dividend

Posted: January 18th, 2017, 11:31 am
by TopOfDaMornin
I was wondering if the more experienced members could suggest factors to look for that would indicate the increase of future dividend.
I personally get the impression that dividend cuts cannot be avoided and diversification across sectors minimises the impact.

For example: Cover > 1.5? Increasing dividend over the last 5 years? Avoid shares where the dividend is too high e.g. 2 times the FTSE 100? Is there evidence to show too high a yield indicates a possible cut?

I am thinking of how to avoid the likes of LLOY, TSCO, SBRY and as off today PSON.
Regards
TopofDaMornin

Re: Factors for future increasing dividend

Posted: January 18th, 2017, 1:16 pm
by OhNoNotimAgain
This is from an investment bank:

FACTORS THAT HELP PREDICT DIVIDENDS CUTS
Poor 12m share price performance
High share price volatility
Poor profitability (low ROE, ROA, high accruals)
Poor balance sheet strength
FACTORS THAT DO NOT HELP PREDICT DIVIDENDS CUTS
Earnings or free cash flow cover
3y, 5y or 10y dividend track record
Historical dividend growth

Rob

Re: Factors for future increasing dividend

Posted: January 18th, 2017, 5:24 pm
by Dod1010
I do not think any of us can avoid a share which cuts its dividend. It will happen to all of us at some time or another but Rob's extract is surely a good guide for all of us, if only as a reminder.

I think a share with a dividend yield of more than 150% of the FTSE100 average is suspect, never mind twice, that is far too high. You can also look at the sector. If it is cyclical, there is a higher risk than say a utility, and there are own goals like Cobham to think of as well. It is not easy and I doubt that any of us will avoid a cutter at some point.

Dod.

Re: Factors for future increasing dividend

Posted: January 18th, 2017, 6:52 pm
by vrdiver
Dod1010 wrote:I do not think any of us can avoid a share which cuts its dividend. It will happen to all of us at some time or another but Rob's extract is surely a good guide for all of us, if only as a reminder.

I think a share with a dividend yield of more than 150% of the FTSE100 average is suspect, never mind twice, that is far too high. You can also look at the sector. If it is cyclical, there is a higher risk than say a utility, and there are own goals like Cobham to think of as well. It is not easy and I doubt that any of us will avoid a cutter at some point.

Dod.



Interestingly, PYAD's metric* for Value (not HYP) shares, was for yield to be preferably 50% above the market. His argument was that whilst waiting for the value to "out" you could at least enjoy a juicy dividend. On the basis that Value and HYP are closely connected** I've always looked for that when selecting my own HYP shares.

VRD


*Source: http://news.fool.co.uk/valueinvesting/1999/vi990820.htm
**Source: http://news.fool.co.uk//valueinvesting/ ... 050930.htm

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 8:04 am
by Wizard
OhNoNotimAgain wrote:This is from an investment bank:

FACTORS THAT HELP PREDICT DIVIDENDS CUTS
Poor 12m share price performance
High share price volatility
Poor profitability (low ROE, ROA, high accruals)
Poor balance sheet strength
FACTORS THAT DO NOT HELP PREDICT DIVIDENDS CUTS
Earnings or free cash flow cover
3y, 5y or 10y dividend track record
Historical dividend growth

Rob

I presume you are suggesting the typical HYP rules for selection of shares are not empirically supported. As I am just starting out on my HYP journey I'd be grateful if you could expand on your list and / or provide more details on the source.

Thanks,
Terry.

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 10:10 am
by OhNoNotimAgain
Wizard wrote:
OhNoNotimAgain wrote:This is from an investment bank:

FACTORS THAT HELP PREDICT DIVIDENDS CUTS
Poor 12m share price performance
High share price volatility
Poor profitability (low ROE, ROA, high accruals)
Poor balance sheet strength
FACTORS THAT DO NOT HELP PREDICT DIVIDENDS CUTS
Earnings or free cash flow cover
3y, 5y or 10y dividend track record
Historical dividend growth

Rob

I presume you are suggesting the typical HYP rules for selection of shares are not empirically supported. As I am just starting out on my HYP journey I'd be grateful if you could expand on your list and / or provide more details on the source.

Thanks,
Terry.


Correct. They were assembled by someone who wanted to provide an alternative to an annuity. However, he did not allow enough for risk and it was done before passive funds became popular. Now you can get a passive fund, that clearly has no alpha risk (i,e, stock specific) and only has the market risk (beta) which is stuff like Brexit. Look around and see what they yield first before attempting to replicate what you can already buy very cheaply.

Rob

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 10:19 am
by funduffer
Dod1010 wrote:I think a share with a dividend yield of more than 150% of the FTSE100 average is suspect, never mind twice, that is far too high.


For I minute I thought that Luni was back......danger zones and all that!

I have to say, the experience in my HYP is that there is definitely something in this. I have had 4 cutters in my 20 share HYP (CNA, SBRY, AMFW, BLT), and another to come (PSON) soon, which will be 25% of the HYP.

I think all of these were at >150% of FTSE yield when purchased.

FD

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 10:59 am
by Plutus
OhNoNotimAgain wrote:...
Correct. They were assembled by someone who wanted to provide an alternative to an annuity. However, he did not allow enough for risk and it was done before passive funds became popular. Now you can get a passive fund, that clearly has no alpha risk (i,e, stock specific) and only has the market risk (beta) which is stuff like Brexit. Look around and see what they yield first before attempting to replicate what you can already buy very cheaply.

Rob


Hello Rob.

Re: passive funds.

I should probably ask on the relevant board, but UKDV is the only passive fund that I'm aware of that could be an alternative to a HYP.

I don't think that e.g. a tracker fund would provide a similar amount of income that a HYP would provide. E.g. FTSE100 yield is approx 3.7% whereas a HYP would aim for 150% above that.

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 12:21 pm
by Dod1010
Well I do not have danger zones but in fact what I was saying was a variation of 'chasing yield' which is not a good idea either. I would not say do not buy if the yield is more than 150% of the FTSE100 average but certainly you need to convince yourself that things have a fair chance of working out. For instance, HSBC and Shell have been around a 6% yield for some time; I think that is because the market thinks there is a reasonable chance that they might not be able to sustain the dividend at current levels. So far that has not been the case but it is something to consider if thinking of buying them.

We have of course moved the discussion away from factors for future increasing dividends to factors for avoiding cuts. Neither is an exact science as we all know only too well and I think that we need to try to focus on factors for avoiding cuts, because that is much more significant to me than worrying about increases in the dividend.

I held Centrica and Cobham and I do not think either were obvious candidates for a cut, and neither I think had a particularly high yield just before the cut. We are really second guessing the Directors which is not often very rewarding because the last things most directors want is to cut the dividend. Fortunately there are usually warning signs as in the case of BLT and Sainsbury at least.

Dod

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 12:30 pm
by OhNoNotimAgain
Plutus wrote:
OhNoNotimAgain wrote:...


I should probably ask on the relevant board, but UKDV is the only passive fund that I'm aware of that could be an alternative to a HYP.

I don't think that e.g. a tracker fund would provide a similar amount of income that a HYP would provide. E.g. FTSE100 yield is approx 3.7% whereas a HYP would aim for 150% above that.


Have a look here
http://www.hl.co.uk/funds/index-tracker ... cker-funds

If you think about it you will see why seeking a 50% yield premium now is almost impossible.

When PYAD set up HYP is was close to the top of the dot.com bubble and there were a lot of expensive stocks with low yields. That is no longer the case.

Rob

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 12:31 pm
by Plutus
OhNoNotimAgain wrote:...

Have a look here
http://www.hl.co.uk/funds/index-tracker ... cker-funds

If you think about it you will see why seeking a 50% yield premium now is almost impossible.

When PYAD set up HYP is was close to the top of the dot.com bubble and there were a lot of expensive stocks with low yields. That is no longer the case.

Rob


Thanks Rob, I have also posed the question on the passive investing sub-forum.

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 7:01 pm
by tjh290633
Plutus wrote:I should probably ask on the relevant board, but UKDV is the only passive fund that I'm aware of that could be an alternative to a HYP.


IUKD is the one that set out to do something similar to an HYP, but it was indiscriminate in its choice of shares, being heavily weighted to finance shares at the time of the 2008 crash.

TJH

Re: Factors for future increasing dividend

Posted: January 19th, 2017, 8:11 pm
by Itsallaguess
funduffer wrote:
For I minute I thought that Luni was back......danger zones and all that!

I have to say, the experience in my HYP is that there is definitely something in this. I have had 4 cutters in my 20 share HYP (CNA, SBRY, AMFW, BLT), and another to come (PSON) soon, which will be 25% of the HYP.

I think all of these were at >150% of FTSE yield when purchased.


I agree that Luni had a reasonably good metric, in a yield over 150% of FTSE yield, to help weed out possible candidates that would merit much closer attention before they were to be 'allowed' near a HYP. It's a sensible approach that has 'one ear' on what the market is trying to highlight in creating such a situation.

Where I think Luni went wrong was to over-complicate what is a really simple idea, and one that can be quickly added to our other 'filters-of-choice' (market-cap / dividend-cover / debt / etc..) in a really simple way, and he chose to develop a full-blown thesis around it, regularly creating thousands of words around the subject.

There never seems a good reason to do a similar 'quantum theory' paper on any market-cap filters, or filters regarding dividend-cover that candidates should satisfy, so I always wondered why it was deemed necessary to do it for comparing a potential HYP-share yield to that of the FTSE yield at any particular time.

Yes, consider it when filtering for potential purchases, as you'd consider any other filter, but I don't think there was ever any need to trouble anyone with any sort of 'Zonal Theory' regarding what is really a very simple concept that can be done at any time on a per-candidate basis.

Cheers,

Itsallaguess

Re: Factors for future increasing dividend

Posted: January 20th, 2017, 8:29 am
by Gengulphus
Itsallaguess wrote:I agree that Luni had a reasonably good metric, in a yield over 150% of FTSE yield, to help weed out possible candidates that would merit much closer attention before they were to be 'allowed' near a HYP. It's a sensible approach that has 'one ear' on what the market is trying to highlight in creating such a situation.

Really?

On November 13th, 2000, the FTSE AllShare yield (which was what Luniversal used, at least when I last paid any detailed attention to his ideas) was 2.19%. That means that any share with a yield of 3.4% or more would be weeded out of a HYP constructed on that day, without any further consideration. (Shares with a yield of precisely 3.3% would be as well, but shares with a yield of 3.3% when quoted to one decimal place might have a precise yield as low as 3.25%, which would pass the test - so I'll give shares with such a yield the benefit of the doubt and assume they're not weeded out.)

Looking at HYP1's initial holdings list on that date in http://news.fool.co.uk//news/foolseyeview/2000/fev001113c.htm, the only holding that would have passed that test is Shell. And looking at the situation 16 years later in viewtopic.php?f=15&t=432&p=4173#p4173, Shell is not exactly a star performer in the portfolio, especially bearing in mind the fact that the holding has been topped up with capital thrown off by other holdings.

I.e. the evidence of HYP1 about filtering for yield < 1.5 * (FTSE AllShare yield) is that it would have thrown out an awful lot of baby with the bathwater - and the little bit of baby it would have kept wasn't exactly a choice bit...

Itsallaguess wrote:Where I think Luni went wrong was to over-complicate what is a really simple idea, and one that can be quickly added to our other 'filters-of-choice' (market-cap / dividend-cover / debt / etc..) in a really simple way, and he chose to develop a full-blown thesis around it, regularly creating thousands of words around the subject.

Where I think he also went wrong was to fail to be critical about his own ideas. Gathering loads of data and seeking out and commenting on the bits that confirm one's theory (such as dividend cutters that the test would have rejected) is not enough: one also needs to seek out the bits that contradict it (such as good dividend performers that it would also have rejected), and to do proper statistical studies - and Luniversal seldom (if ever) showed evidence of having done such things.

And even if there is something useful in the general idea, is the FTSE AllShare yield the best index yield to use? Is 1.5 the correct multiplier to use? Should it be a multiplier at all, or would e.g. an addition be better? (*) Again, I saw little or no evidence of having explored such questions.

(*) For instance, a test for the share's yield being under the FTSE AllShare yield plus 2 percentage points would have produced broadly similar results much of the time - for FTSE AllShare yields in the 3-4% range, it's a mildly less strict filter than yield < 1.5 * (FTSE AllShare yield). But for HYP1, constructed when the FTSE AllShare yield was abnormally low, it would have been quite a bit less strict, letting the five original companies with yields <= 4.2% through - and those companies do at least include one that went on to perform well, namely Associated British Ports.

Itsallaguess wrote:There never seems a good reason to do a similar 'quantum theory' paper on any market-cap filters, or filters regarding dividend-cover that candidates should satisfy, so I always wondered why it was deemed necessary to do it for comparing a potential HYP-share yield to that of the FTSE yield at any particular time.

There's good reason to try to study all of them - but studying them properly requires a large investment of time and effort and careful design of the study to avoid hindsight bias and confirmation bias. Few private investors can afford that investment, far fewer are willing to prioritise it over other uses of the time and effort, and even fewer have the detailed understanding of the various pitfalls required to do the design properly. And the net result is that we don't have even one proper "'quantum theory' paper" about Luniversal's yield < 1.5 * (FTSE AllShare yield) test, let alone similar ones about other tests.

Gengulphus

Re: Factors for future increasing dividend

Posted: January 20th, 2017, 8:40 am
by Wizard
Gengulphus wrote:
Itsallaguess wrote:I agree that Luni had a reasonably good metric, in a yield over 150% of FTSE yield, to help weed out possible candidates that would merit much closer attention before they were to be 'allowed' near a HYP.


That means that any share with a yield of 3.4% or more would be weeded out of a HYP constructed on that day, without any further consideration.

my emphasis

A little unfair to say the least to quote Itsallaguess and then criticise something he didn't actually say :roll:

Terry.

Re: Factors for future increasing dividend

Posted: January 20th, 2017, 11:54 am
by Gengulphus
Wizard wrote:
Gengulphus wrote:
Itsallaguess wrote:I agree that Luni had a reasonably good metric, in a yield over 150% of FTSE yield, to help weed out possible candidates that would merit much closer attention before they were to be 'allowed' near a HYP.


That means that any share with a yield of 3.4% or more would be weeded out of a HYP constructed on that day, without any further consideration.

my emphasis

A little unfair to say the least to quote Itsallaguess and then criticise something he didn't actually say :roll:

Yes, there's an ambiguity there - basically, whether "to help weed out possible candidates ..." was part of the description of what Luni had, or whether it had finished saying what Luni had ("a reasonably good metric, in a yield over 150% of FTSE yield") and was proceeding to say what HYPers could do with that metric. I read it the first way and applied my knowledge that Luni generally just weeded out the shares concerned, seldom giving them any further consideration. But on re-reading it, the second way does make more sense and was probably what Itsallaguess intended - so sorry about reading it the way I did!

Having said that, it still looks to me like a not-very-useful metric in the case of HYP1, because it would have been just about as easy to give all 15 shares "much closer attention" as to use their yields to decide that one of them didn't need it. And Shell's subsequent history is not so unblemished that I would be happy in retrospect not to have given its selection the same amount of attention as the rest...

Gengulphus

Re: Factors for future increasing dividend

Posted: January 20th, 2017, 12:41 pm
by Itsallaguess
Gengulphus wrote:
Yes, there's an ambiguity there - basically, whether "to help weed out possible candidates ..." was part of the description of what Luni had, or whether it had finished saying what Luni had ("a reasonably good metric, in a yield over 150% of FTSE yield") and was proceeding to say what HYPers could do with that metric. I read it the first way and applied my knowledge that Luni generally just weeded out the shares concerned, seldom giving them any further consideration. But on re-reading it, the second way does make more sense and was probably what Itsallaguess intended - so sorry about reading it the way I did!


All I can say is that I'm glad I returned to this thread after the above re-consideration, rather than the earlier crucifixion! :lol:

Yes, what I clearly meant was that I think it's a fairly good metric to help highlight possible candidates that could merit a deeper analysis, over and above the possibly less-onerous analysis that we might choose to carry out over shares that are not yielding over 150% of the current FTSE yield.

That's of course a deeper analysis because we should be un-trusting of such relatively high-yields, and we'd be looking for more evidence that purchasing such a share would be a good thing to do, not a deeper analysis because the metric might tell us it should instantly become a favourite-choice of any sort simply on the back of the metric....

I wouldn't ever think that such a metric should be used carte-blanch to remove possible candidates completely from consideration, and I'm glad that if there was any confusion over that position, then it's hopefully now been cleared up. Thanks Wizard for helping to point out my intended meaning.

Cheers,

Itsallaguess

Re: Factors for future increasing dividend

Posted: January 20th, 2017, 1:05 pm
by Itsallaguess
Gengulphus wrote:
Itsallaguess wrote:There never seems a good reason to do a similar 'quantum theory' paper on any market-cap filters, or filters regarding dividend-cover that candidates should satisfy, so I always wondered why it was deemed necessary to do it for comparing a potential HYP-share yield to that of the FTSE yield at any particular time.


There's good reason to try to study all of them - but studying them properly requires a large investment of time and effort and careful design of the study to avoid hindsight bias and confirmation bias. Few private investors can afford that investment, far fewer are willing to prioritise it over other uses of the time and effort, and even fewer have the detailed understanding of the various pitfalls required to do the design properly. And the net result is that we don't have even one proper "'quantum theory' paper" about Luniversal's yield < 1.5 * (FTSE AllShare yield) test, let alone similar ones about other tests.

Gengulphus


Completely agree with all of the above, but I think it's over-complicating my position on the topic.

I think that even if empirical evidence existed today that there was some 1xx% FTSE-yield figure that could be proven to foresee relatively high danger with regards to the safety or predictability of future dividend-income from potential HYP purchases, I still can't see the justification for the whole 'Zonal' theory, with its accompanying words and threads over the years. All that stuff is/would be using a hammer to crack a nut at that point, as far as I'm concerned.

Why don't we have a 'Zonal' theory for dividend-cover? Anyone who looks at dividend-cover, and I do myself sometimes, might want to see a cover of over 1 and might begin to sweat a little over an ultra-high yielding candidate with a cover of less than 0.9, let's say by way of an example.

So in a way we might look to see that 0.9 figure, for the sake of this discussion, to represent the magic 150% FTSE-yield figure of this yield metric. Why then, don't those who look at dividend-cover as this type of filter need to also use some sort of 'Dividend-cover Zonal Theory' to go along with that test?

They don't need it because they can look at it instantly, on a per-candidate basis, and make a judgement in seconds as to the relevance of it.

They don't need to spend hours and hours compiling Zonal lists of FTSE shares and their Optimum/Danger/etc. Zones that their specific dividend-cover-level is sitting in; they can simply look at the figure, in exactly the same way that we can look at a current or forecast-yield figure, and make a judgement on a case by case basis.

I don't need to know that Royal Mail might be in the Yield-Danger-Zone if Royal Mail is never going to be a potential candidate for my HYP purchase, but if Carrillion pop up on my candidate list, then I can take the whole five seconds it takes, if it's important to me, to view the current or forecast-yield of Carrillion against the current yield of the FTSE, and make a judgement in those five seconds whether I need to carry out more due-diligence, or whether I want to simply dismiss the candidate from the list.

So my issue wasn't ever really with the lack of real evidence that the Zonal theory worked, or the quality of the 'evidence' that did exist, although I agree that those issues are clearly very important; my issue was that even if someone did think it worked, then they don't really need to rely on someone coming up with a regularly updated list of shares, along with their Yield Zones, to use it....

The whole thing created a dependency that was, in my view, wholly unwarranted. It was the antithesis of everything the Fool used to stand for, in terms of helping people to help themselves....

Cheers,

Itsallaguess

Re: Factors for future increasing dividend

Posted: January 21st, 2017, 9:25 am
by Gengulphus
Itsallaguess wrote:Yes, what I clearly meant was that I think it's a fairly good metric to help highlight possible candidates that could merit a deeper analysis, over and above the possibly less-onerous analysis that we might choose to carry out over shares that are not yielding over 150% of the current FTSE yield.

That's of course a deeper analysis because we should be un-trusting of such relatively high-yields, ...

I think that's somewhat over-complicating it. IMHO we should be un-trusting of all high yields - that's why the HYP approach requires checks / analysis about dividend safety. Skimping on those checks / analysis because the yield is under 150% of the current FTSE yield seems rather dangerous to me - and HYP1 backs that up: it would only have saved one lot of deeper checks / analysis out of 15, and given that Shell had its reserves problems emerge a few years later, it's by no means clear to me that saving that one lot of them would really have been desirable!

Personally, I would keep things simple: have a decent set of checks / analysis that one applies to all candidates. After doing them, one either decides "safe enough" or "too dangerous".

Gengulphus

Re: Factors for future increasing dividend

Posted: January 21st, 2017, 9:35 am
by Raptor
Gengulphus wrote:
Personally, I would keep things simple: have a decent set of checks / analysis that one applies to all candidates. After doing them, one either decides "safe enough" or "too dangerous".

Gengulphus


Well put 8-) . All I would add is that as you go from the "building phase" to the "taking income phase" keep an open mind to "changes" to your "checks and balances". Strange how your "mindset" changes as you start to look at taking income........

Raptor.