Forecast Yields - do we believe them yet?

[Arborbridge]

Why do investors need a forecast dividend anyway? What is the purpose? In normal times I use the trailing yield (actual income for the last 12 months/ current value) but currently that does not work for all shares because as we know same have cancelled their dividends and others have drastically reduced them. For me that applies only to Imperial Brands and Shell having reduced their dividends and to HSBC having cancelled them. Otherwise for the sake of comparison, I can continue to use the trailing yield.

That avoids my having to worry about forecasts and how to check them and avoids worrying about the reliability of any particular site. Keep it simple.

Dod

Why would I work out a trailing yield, when a forecast yield automatically pops out of the HYPTUSS? And it seems you have some "worry" too as you admit that doe not work!

Why would I go to that effort? Keep it simple, indeed - run HYPTUSS. The results without correction or faffing about is likely to be as close as the trailing yield (in point of fact, HYPTUSS will return the trailing yield too, if that's what you want).

As regards the purpose - why do you want the yield?

For me it's twofold: when buying new shares I need to know the likely outlook (so clearly the trailing yield is not relevant), for the existing invesments I want a reasonabl idea of my income, for I usually set that each December to run for a year. 2020 was the first time I reduced my income part way through the year, but hopefully that will be an exception. Luckily, it was one I planned for.

Arb.

[Arborbridge]

I do try to estimate dividend income approximately for the tax year, but mainly so I can ensure investments are switched between Mrs Scrumpy and me to optimise tax allowances. other than that, given that anyone relying on dividends for spending should have set up a substantial cash buffer, I too can't see the point of obsessing about forecast dividends. What will be, will be.

Similarly I can't see the point of trying to work out XIRRs of previous years investment returns. It doesn't change anything and as we are often told 'the past is no guide to the future', though perhaps it is with SMT?

XIRR is an interesting one. I use it to keep an eye on investments to see the relative ebb and flow. If one is a permanent "bad boy" over many six month segments (amounting to years) I would consider moving on, but wouldn't use XIRR alone to decide. This comment BTW, is mainly applicable to ITs.

I too can't see the point of obsessing about forecast dividends. What will be, will be.

Obsessing? Never but for planning or decision making, and most sensible people would base both on information. You can fly blind if you like, but I prefer some visiblity even if somewhat foggy (as Jim Slater once wrote to me)

Arb.

[Dod101]

That avoids my having to worry about forecasts and how to check them and avoids worrying about the reliability of any particular site. Keep it simple.

Keep it simple to me is looking forwards not backwards, but we are all different in how we approach these things I guess.

I like to base my decisions on fact not forecast, and I assume that dividends will at least be maintained, unless we have been told otherwise. That style has served me well over the last 25 years or so, and has often brought a pleasant surprise with it.

In fact the only real shock I have had is when HSBC was forced by the PRA to cancel its dividend within a very few days of its being paid in March/April this year.

Dod

[Itsallaguess]

I like to base my decisions on fact not forecast, and I assume that dividends will at least be maintained, unless we have been told otherwise.

So suddenly you're not basing your decisions on 'facts' at all, but 'assumptions', and another word for an 'assumption' might be a 'forecast'....

I wonder how the 'assumptions' made 12 months ago for the forthcoming year actually played out for many holdings...

This is a terminology issue, and you're a very good example of someone who's convinced themselves that 'assumptions' are 'facts', and not 'forecasts' Dod...

Cheers,

Itsallaguess

[seagles]

To be honest I only really look at "forecast yields" when I am looking at "adding" to my HYP a new share (or looking at whether better to buy something new or just top-up). Other than that I am more interested in current yield or already announced future yields for "who to top-up" (which has not been a problem since March ) or what to expect as income for next year.

[Dod101]

[
So suddenly you're not basing your decisions on 'facts' at all, but 'assumptions', and another word for an 'assumption' might be a 'forecast'....

I wonder how the 'assumptions' made 12 months ago for the forthcoming year played out for many holdings...

This is a terminology issue, and you're a very good example of someone who's convinced themselves that 'assumptions' are 'facts', and not 'forecasts' Dod...

I find all this quite good fun. My facts are what has been paid in the last 12 months. If I do not know otherwise, I simply assume that the dividends will not be reduced. We know that Shell has reduced its dividend and also that Imperial Brands has so I am aware that their historic yield is not going to be repeated but for my other holdings I simply have to assume that the dividends will at least be held. An assumption is not a forecast. To do a forecast I would have to sit down and consider each share individually and decide how much I thought they would pay as a dividend. I have no idea if HYPTUSS does that or not, but I would never do that. So then I say, so long as the overall dividend level is being maintained, that is fine. Until this year, it has usually increased a bit which is even better, but I still come back to the point, why stress on forecasts? What are you going to do with them anyway?

When first HSBC stopped its dividend and then Shell and Imperial cut theirs, I decided to cut my expenditure by 25%. In the even it is more like a 33% cut, involuntary on my part, because I am not eating out much, nor doing much travelling nor much else. We now approach the year end and forecasters will need to get out their crystal balls I assume, in order to forecast an income from dividends for next year. My method does not need that because I just assume that my dividend income will be at least as much as this year. In fact, it will probably be more if HSBC restores some sort of dividend. Would you forecasters like to give an indication of your forecast for HSBC's restored dividend, if any? The range of possibilities is far too wide to make any forecast and the only sensible one is nil, that is the same as for 2020. We may get a bit more and my method will treat that as a nice bonus but I do not fuss over it as I assume a forecaster must.

Dod

[Itsallaguess]

Would you forecasters like to give an indication of your forecast for HSBC's restored dividend, if any? The range of possibilities is far too wide to make any forecast and the only sensible one is nil, that is the same as for 2020.

We may get a bit more and my method will treat that as a nice bonus but I do not fuss over it as I assume a forecaster must.

Sharecast is currently showing a Forecast Yield for HSBA of 1.7% -

https://www.sharecast.com/equity/HSBC_Holdings

There's no 'fuss' in getting that data - it's all done automatically for many of us who use that facility, so overplaying the 'fuss' card is actually another red-herring Dod - there's no 'fuss' in using these broker-forecasts, as the brokers do all the 'fussing' for us as part of their day-job in talking to the companies themselves, and it's clear from hearing how the 'non-forecasters' are doing things on this thread that there's actually a lot more 'fuss' in those methods in terms of manual-interventions and data-gathering than using this freely available broker-forecast data...

Anyway, I think it's a good opportunity to test our different theories on something like HSBA, so at a current price of 337.8p, and a Sharecast Forward Yield of 1.7%, they are proposing a total dividend in the next 12-month period of around 5.74p, whereas you're manually assuming there will be no dividends paid in the next 12 months, so it'll be an interesting exercise to re-visit this thread later next year.

Please though - using the word 'fuss' numerous times when discussing these Sharecast forecast yields, at the same time as admitting that you don't know how people are getting them really isn't helping the discussion...

Cheers,

Itsallaguess

[dealtn]

Would you forecasters like to give an indication of your forecast for HSBC's restored dividend, if any?

Not in this instance, as I don't hold it, or consider buying it so I haven't done the work. Were it to move close to being an investment choice I would be making some kind of forecast which I might share. I can't see that happening anytime soon though.

[Dod101]

Well if HSBC is to pay any dividend I think 5.74p per share is pretty safe but it is also meaningless. That would be a massive drop on where they were in 2019 and were I to believe that then I think I would be selling my HSBC shares now. We will not I guess get a full reinstatement but to have any credibility it will need to be a lot more than that. They were paying 51 cents in the year before the cancellation.

I am assuming no dividend but that is not the same as not expecting a dividend. If the PRA agrees, then I think we will see more than that over the course of the year. In fact that sounds like a quarterly payment if my exchange rate is anything like right.

Dod

[Itsallaguess]

Well if HSBC is to pay any dividend I think 5.74p per share is pretty safe but it is also meaningless.

Not in relation to this specific discussion it's not...

Cheers,

Itsallaguess

[moorfield]

A forecast yield or income in itself is not particularly useful. A forecast versus target portfolio income on the other hand, is, if one is extrapolating further into the future. Against a future target I can measure progress and (roughly) discount to present value. I can also size my cash float in terms of "months of target income" (currently about 8) which is a useful way of thinking, and this year particularly has helped me figure out what the minimum float I should aim to always be holding is before I have to start selling the furniture (about 4). That my forecast "ntm" (next twelve months) income (whether it uses trailing dividends, those announced otherwise, or a pro-rata combination of both) will be well well below target at the end of this year tells me I will need to contribute more cash or upcycle low yielders in my portfolio next year.

[dealtn]

Well if HSBC is to pay any dividend I think 5.74p per share is pretty safe but it is also meaningless. That would be a massive drop on where they were in 2019 and were I to believe that then I think I would be selling my HSBC shares now. We will not I guess get a full reinstatement but to have any credibility it will need to be a lot more than that. They were paying 51 cents in the year before the cancellation.

I am assuming no dividend but that is not the same as not expecting a dividend. If the PRA agrees, then I think we will see more than that over the course of the year. In fact that sounds like a quarterly payment if my exchange rate is anything like right.

Dod

How can you make a decision about selling now, concurrently believing 5,74p to be too low, and worthy of a sell decision, without having a view (or forecast) on what the dividend is to be? I can't "square that circle" in your thinking about this.

[funduffer]

I am with IAAG and Arb on this.

I only use yield forecasts through HYPTUSS in selecting candidates for top-ups, and periodically for getting an estimate of HYP total income for the year ahead for wider financial planning.

Of course once I have a top-up candidate I will do further research to try and gauge the sustainability of the dividend.

My point is that back in April the forecast yields would be obviously wrong, but I suspect they are now much more reliable, for these purposes.

It's no fuss to run HYPTUSS, just press the 'update prices' and 'forecast yields' buttons. Kiloran and IAAG have done all the work for me!

FD

[Dod101]

Two things. It may be no fuss to run HYPTUSS but what it tells you is surely only as good as how reliable the input is.

As for dealtn's comment, I am taking the forecast from IAAG and if it is anything like correct, it would not be worth hanging on for the dividend resumption. I am prepared to think that it may well be more than that, so will hang on. The fact is though that one man's guess is as good as another's. That is why forecast yields especially when thinking of a dividend resumption, are so pointless.

Dod

[Dod101]

I think I am correct in saying that at the Q3 results announcement the HSBC Finance Director said that if the dividend resumed for the Final in 2020 it would not be a token and that is what I think the so called 'forecast' is, so I think it will either be more than that or nothing at all. Thus to my mind, this 'forecast' is worse than useless. That is why I would prefer just to assume that the future dividends will be not less than the historical ones. If they turn out to be more, then good and well.

Dod

[Itsallaguess]

The fact is though that one man's guess is as good as another's...

But that's simply not true, and we have the evidence on this thread to prove it....

When the data-source we're talking about has been used for many years by a number of different people, who are happy to tell you on this thread that they've found those particular forecasts reliable enough over long enough periods to continue to be useful to them, we're suddenly not in the same realm as 'one man's guess being as good as another's', because we can begin to apply a certain evidence-led pedigree to one of those particular data-sources..

You've not used it, which is absolutely fine, but I'm very surprised to see a level of dismissal from you that goes well beyond that, and seems to also want to dismiss other people's experience of it too...

Of course I wouldn't ever dare suggest for a minute that each and every forecast might be 100% accurate - who would? - but I can tell you that I've found enough of them accurate, and for long enough periods, for them to be very useful to me for many aspects of my income-investment strategy...

If I tell you that for a number of years before this very odd and COVID-induced outlier one, the forecasts of my overall portfolio's yearly future income using this data-source has been accurate to within a very small amount once that income had actually come in, then why would I ever want to think of that data-source as being 'only as good as any other man's guess'?

Cheers,

Itsallaguess

[Dod101]

IAAG. When I used the expression that' one man's guess is as good as another' I was thinking about HSBC in particular but I have probably said enough on this topic. I am sure you get my message which in itself may be a load of cobblers but time will tell.

Dod

[Arborbridge]

I think I am correct in saying that at the Q3 results announcement the HSBC Finance Director said that if the dividend resumed for the Final in 2020 it would not be a token and that is what I think the so called 'forecast' is, so I think it will either be more than that or nothing at all. Thus to my mind, this 'forecast' is worse than useless. That is why I would prefer just to assume that the future dividends will be not less than the historical ones. If they turn out to be more, then good and well.

Dod

I went even further and assumed a zero dividend! As you say If they turn out to be more, then good and well.
Whatever the dividend is likely to be it a) will not be life-changing and b) won't affct whether HSBC is in the topup zone as it is currently way below the radar in 24th.

Arb.

[kempiejon]

Since Covid struck in March, I have done precisely nothing to my HYP, mainly because I realised that forecast yields (the underlying parameter in...

My conclusion is to start taking forecast yields more seriously again, as they look reasonable to me - at least in my HYP.

FD

Thanks for running the numbers and sense check on the data. I've been updating my SS and looking at news and statements and had by my ready reckoning come to the same conclusions regarding the veracity of the Sharecast forecasts and hence HYPTUSS numbers at portfolio level.

[tjh290633]

... I use the latest available data, as issued by the company concerned. ...

Exactly what do you mean by that? Data about dividends that have actually been paid? Data about dividends the company has declared, but not yet paid? Data about dividends the company has said it intends to declare, but has not yet actually declared? They're all data issued by the company, but the dividends those data are about have varying degrees of certainty of actually ending up being paid (only the first category has a 100% chance, as e.g. the banks demonstrated early this year), various amounts of useful detail (e.g. DS Smith's statement that it intends to pay an interim dividend this year is not very useful because it doesn't say a thing about the size of that dividend), and various other things affecting its reliability and usefulness...

I use historical data, i.e. dividends that have been declared and/or paid by the company. Unless the company concerned cancels its dividend, it stays as declared.

There are one or two assumptions in there, like the level of the dividends of IMB to be announced on the 17th. ...

Is that an example of an irregular verb phrase? - something like "I make assumptions, you make forecasts, he makes wild guesses"? ;-)

I quote from the RNS https://www.investegate.co.uk/imperial- ... 00023186N/ giving the half year results for IMB:

Dividend to be rebased by one third; implying an annual dividend for 2020 of 137.7 pence per share

I am assuming that that will be the level of the total dividends for 2020, to be announced on the 17th, as I said. It is the best guidance that I have at the moment.

In the case of DS Smith, my current assumption is for zero dividend until I know otherwise. We shall know on 10th December what they are going to pay.

TJH