Dividends in the next FY?

[monabri]

Anything above RPI would be nice but even nicer would be " no more Carillion, Interserve nor PFGs". A swing to the positive. <Pendant mode off>

[Arborbridge]

My main concern is blowing the Standard Rate band. I have done some careful calculations for the next FY. I have decided to go with an 8% dividend increase, which matches the forecast for the FTSE 100, and is reasonably consistent with last year's increase. I suspect that is optimistic, but looks as though I will finish just under the wire. Nonetheless, the Higher Rate taxman is stalking me. I will make some adjustments to give myself more breathing space, when I have the cash flow.

Why? We are buying a sub-set of the FTSE100 (and some of us buy outside the 100) so the increase of that sub-set may be different. Indeed, if one believes the old wives' tales and Luni, higher yield may lead to lower increases and vice-versa.
Instead of relying on some general market wide increase which is not appropriate, one could have put one's HYP into HYPTUSS and examine that subset of shares separately.


Arb.

[Arborbridge]

Anything above RPI would be nice but even nicer would be " no more Carillion, Interserve nor PFGs". A swing to the positive. <Pendant mode off>

You're about right there! It's often said that one's failures are more important and the decline more precipitous.

[GeoffF100]

Why? We are buying a sub-set of the FTSE100 (and some of us buy outside the 100) so the increase of that sub-set may be different. Indeed, if one believes the old wives' tales and Luni, higher yield may lead to lower increases and vice-versa.
Instead of relying on some general market wide increase which is not appropriate, one could have put one's HYP into HYPTUSS and examine that subset of shares separately.

I have been glossing over some details here. What I really want is an estimate of the growth for my equity holdings that are outside tax shelters. This includes an 18 year old UK portfolio that passes for a HYP, but is no longer particularly high yield. It also includes overseas tracker funds. The trackers have the majority of the capital, but produce only about a third of the dividends. I am coppicing the HYP to fill my CGT allowance and add to my tracker holdings. I have also been adding some fresh capital to the trackers. I have done some guesstimates to get to the 8%.

I only need a rough idea here. A taxable dividend growth of just over 10% would push me out of the Standard Rate band, but a charitable contribution of a few thousandths of a percent of my portfolio value should be enough fix that. (You can make a charitable donation for the current tax year up to 31 January in the next tax year, and declare it on your tax return, provided that you file it by 31 January. You cannot add a charitable donation by changing a tax return that you have already submitted though.)

[Gengulphus]

My main concern is blowing the Standard Rate band. I have done some careful calculations for the next FY. I have decided to go with an 8% dividend increase, which matches the forecast for the FTSE 100, and is reasonably consistent with last year's increase. I suspect that is optimistic, but looks as though I will finish just under the wire. Nonetheless, the Higher Rate taxman is stalking me. I will make some adjustments to give myself more breathing space, when I have the cash flow.

It's dead easy to avoid "blowing the Standard Rate band" if that's your highest priority - just make sufficient pension contributions and/or Gift-Aided donations to charity to expand the band by enough to cover all your income. The amount you can do with pension contributions is limited, but the amount that you can do with Gift-Aided contributions to charity isn't...

It's not so easy if you have higher priorities than avoiding "blowing the Standard Rate band", such as maximising the amount of spending money you have available. The Higher Rate taxman is not a dangerous stalker on his own, but he becomes one if he finds suitable allies among the taxpayer's priorities! ;-)

Edit: Sorry, I failed to look ahead in the thread and see your subsequent comments about a modest charitable donation, which removes part of the point of this post of mine. But there's still the point that it needn't be a modest one, and what I at least regard as some mild humour about the Higher Rate taxman - hence this edit rather than just deleting my post as redundant.

Gengulphus

[GeoffF100]

You are right Gengulphus, my charitable donation need not be modest. My life expectancy is probably not more than twenty years now, but given my excellent health I may beat the odds. Nonetheless, it is not remotely likely that I will ever spend all the money that I have accumulated since retiring nearly twenty years ago. There is a strong argument for giving some money away before I finally have no choice. There is a saying that money is a good servant and a bad master, and perhaps I should pay heed to it.

[Julian]

I'm less optimistic than some about the level of my dividend growth from my locked-down HYP (i.e. no new capital going in) next FY. I will be disappointed and slightly surprised if my total dividend income actually drops, I will be fairly OK about it and not wringing my hands in angst if my dividend income stays at the same level next FY, and I'll be actively happy if I were to see a 1% increase next FY. I would be very happy indeed to see anything more than 1% growth.

The reason for my pessimism is exemplified by today's BP announcement but there are many other examples from Shell, HSBC etc where Sterling's recovery has, in the case of BP today for instance, resulted in a reduction of more that 12% in the Sterling divi announced today vs the one last March. These currency effects are very definite steps back in my portfolio income that are offsetting the nice gains from the likes of BATS. I do note that today's BP divi is in this FY but I am concerned that these effects of Sterling's unwinding weakness might continue into next year.

The counter is that currency effects go both ways so who knows whether these USD and EUR declarers will start going the other way again if Sterling weakens but right now it just feels to me, purely emotionally and with no analysis, that next year's journey into dividend growth will be done in the face of somewhat of a headwind still coming from exchange rate effects.

- Julian

[tjh290633]

The reason for my pessimism is exemplified by today's BP announcement but there are many other examples from Shell, HSBC etc where Sterling's recovery has, in the case of BP today for instance, resulted in a reduction of more that 12% in the Sterling divi announced today vs the one last March. These currency effects are very definite steps back in my portfolio income that are offsetting the nice gains from the likes of BATS. I do note that today's BP divi is in this FY but I am concerned that these effects of Sterling's unwinding weakness might continue into next year.

Although the change from last year's March dividend is considerable, that is down to the exchange rate for the two dividends. For 2017 as a whole, the fall is only 1.89% on an unchanged US$ dividend converted to GB£ at the quoted exchange rates which were:

Date         USc       pence    Rates  
19-Jun-15    10.000    6.530    1.5315 
18-Sep-15    10.000    6.549    1.5270 
18-Dec-15    10.000    6.634    1.5073 
24-Mar-16    10.000    7.013    1.4260 
17-Jun-16    10.000    6.917    1.4458 
16-Sep-16    10.000    7.558    1.3231 
16-Dec-16    10.000    7.931    1.2608 
31-Mar-17    10.000    8.159    1.2257 
23-Jun-17    10.000    7.756    1.2893 
22-Sep-17    10.000    7.621    1.3121 
21-Dec-17    10.000    7.444    1.3435 
29-Mar-18    10.000    7.169    1.3949

What we are looking for is an increase in the US$ dividends for 2018. Currency fluctuations will always be with you. Are you complaining that the Pound is stronger now than it was a year ago?

TJH

[Arborbridge]

Currency fluctuations will always be with you. Are you complaining that the Pound is stronger now than it was a year ago?

TJH

It's always swings and roundabouts. Julian doesn't look as though he is complaining, in my view: just giving an opinion about the likely outcome.

I agree with him. There's nothing currently in the forecast delivered by HYPTUSS which suggests a decent increase: something between 0 and 4% seems probable. Neither is there much salvation in capital increases to cheer us up since the nation shot itself in the foot.


Arb.

[Arborbridge]

I agree with him. There's nothing currently in the forecast delivered by HYPTUSS which suggests a decent increase: something between 0 and 4% seems probable. Neither is there much salvation in capital increases to cheer us up since the nation shot itself in the foot.


Arb.

Just to put some flesh on that, here's a chart showing the variation in forecast dividends per unit in pence. Each point is the forecast for the next 12 months from HYPTUSS. One would not expect it to be exact, but I'm interested in both the general prediction and the change against time. The yellow line is the HYP units price in pence, left hand scale.

Given that my dividends amounted to 6.45p for 2017, you will see that the forecast for 2018 made at the end of 2017 was slightly negative, with some improvement since then. That's why I'm currently saying between 0 and 4% - naturally it can change!
It's also clear that the capital has been suffering since the referendum implications are being gradually realised by analysts, and also with political pressure on utilities plus the CLLN disaster more recently.

[Gengulphus]

I agree with him. There's nothing currently in the forecast delivered by HYPTUSS which suggests a decent increase: something between 0 and 4% seems probable. Neither is there much salvation in capital increases to cheer us up since the nation shot itself in the foot.

Just to put some flesh on that, here's a chart showing the variation in forecast dividends per unit in pence. Each point is the forecast for the next 12 months from HYPTUSS. One would not expect it to be exact, but I'm interested in both the general prediction and the change against time. The yellow line is the HYP units price in pence, left hand scale.
...

I'm afraid I find the variations in dividends per unit that chart shows rather hard to believe, in particular with regard to some of the jumps it shows. The clearest example is that it appears to jump up by about 0.25-0.3 out of around 7 in January 2017 and then down again by the same amount or a bit more in March 2017. That's about 3.5-4%, or roughly an entire holding's worth of dividends in a 25-30 share HYP, which is really a very big change to happen as abruptly as those jumps... A sharp jump downwards of that size is comprehensible if a holding cancels its dividend, though even then it seems a bit questionable given that most cancelled dividends hit actual payments in two half-yearly or 4 quarterly instalments, and forecasts tend to take some time to adjust. But I can't think of any likely scenario that explains a share jump upwards of that size - and in addition I haven't managed to think of any dividend cancellation that would have hit in March 2017 (though that may of course just be a memory failure on my part).

So could I ask you to look at your data and your calculations from it to see what did cause those two jumps?

Gengulphus

[Tula100]

My main concern is blowing the Standard Rate band. I have done some careful calculations for the next FY. I have decided to go with an 8% dividend increase, which matches the forecast for the FTSE 100, and is reasonably consistent with last year's increase. I suspect that is optimistic, but looks as though I will finish just under the wire. Nonetheless, the Higher Rate taxman is stalking me. I will make some adjustments to give myself more breathing space, when I have the cash flow.

Why are these investments not in an ISA?

[Gengulphus]

Why are these investments not in an ISA?

Earlier in the thread:

I am taking full advantage of ISAs and a SIPP.

Hopefully that should indicate the answer to your question!

Gengulphus

[GeoffF100]

Why are these investments not in an ISA?

Because the allowances have always been way too small.

[Arborbridge]

[
So could I ask you to look at your data and your calculations from it to see what did cause those two jumps?

Gengulphus

Yes, I could have a look, and of course I was also alarmed at the time, though could find no obvious cause at the time. However, it doesn't particularly worry me as it is the general picture I am after. Some of the explanation could not easily be traced now, and that is if DL had some errors in yields. Not also that these are percentages which come straight from the HYPTUSS without any attempt to check or correct for bloomers which do occur from time to time in DL.

Note that the "raw" uncorrected look ahead in December 2016 gave an income of around the 6.7p mark, but I only achieved 6.45 by Dec 2017. The income predicted now is not so far different, unfortunately.


Arb.

[GeoffF100]

Here is a helpful article:

https://www.ajbell.co.uk/news/ftse-100- ... d-tempting

The forecast increases in FTSE 100 dividends are concentrated in a few shares and few sectors, particularly financials.

[Julian]

Currency fluctuations will always be with you. Are you complaining that the Pound is stronger now than it was a year ago?

TJH

It's always swings and roundabouts. Julian doesn't look as though he is complaining, in my view: just giving an opinion about the likely outcome.

I agree with him. There's nothing currently in the forecast delivered by HYPTUSS which suggests a decent increase: something between 0 and 4% seems probable. Neither is there much salvation in capital increases to cheer us up since the nation shot itself in the foot.


Arb.

Indeed. I was going to object to the term "complaining" myself, I was simply making an observation. If I was to restrict myself strictly to the topic of this thread, i.e. the prospects for next FY's dividends from someone's (in my case my) HYP, I suppose you could say that I was complaining about it but maybe not even then since the USD-GBP exchange rate has implications for companies, some of them potentially positive, far beyond what Sterling equivalence rate the various USD and EUR declarers might use.

What we are looking for is an increase in the US$ dividends for 2018.

Agreed. There is one other observation I would make that relates to that which was in the back of my mind when I made my last post but I didn't type it out at the time...

There seems to have been some correlation, at least in my portfolio, between USD declarers and flat native (i.e. USD) dividends, e.g. HSBC on a $0.10/$0.10/$0.10/$0.21 cycle for the last 3 years and similar "stuck" dividends for Shell ($0.47 every quarter for the last 4 years), possibly BP but I can't be bothered to do the research on that one and maybe others I have forgotten. The combination of a USD-divi-declarer that has also kept its USD divis frozen for a few years are the ones that I really feel contribute to the "headwinds" for dividend growth next year and yes, maybe those do cause me to grumble just a bit.

I'm not saying these headwinds mean that my dividend income is doomed to drop next FY, I'm just saying that unless things change (exchange rate or USD declaration amounts) then my other HYP constituents will have to work harder to give me the growth I hope for. Being a fairly conservative forecaster my back-of-my-mind expectation (I'm deliberately not calling it a forecast) is simply for flat income next FY vs this FY.

Thanks Arb for the charting detail by the way.

- Julian

[Gengulphus]

Yes, I could have a look, and of course I was also alarmed at the time, though could find no obvious cause at the time. However, it doesn't particularly worry me as it is the general picture I am after. Some of the explanation could not easily be traced now, and that is if DL had some errors in yields. Not also that these are percentages which come straight from the HYPTUSS without any attempt to check or correct for bloomers which do occur from time to time in DL.

Ah, I'd missed the fact that you specifically said it's a chart of forecast income per unit. Can't say that there's anything clearly wrong about doing so, but unitisation is basically a way to measure actual performance and so my mindset when looking at the chart was on that... So apologies, I should have read your text more carefully.

Not certain what all the effects are that using forecast dividends could have, but I suspect there are quite a few, many of them not actually associated with the company. Nor necessarily DL, for that matter (though it certainly is at least sometimes, e.g. when they've failed to take an exchange rate into account). For instance, if a particularly optimistic or pessimistic forecaster ceased to contribute to the consensus forecasts for a couple of months, it would push DL's consensus forecasts down or up respectively for those months without DL doing anything wrong. And I strongly suspect that forecasters aren't immune to the prevailing mood of the market, whether that mood is justified or not - e.g. when updating the forecast yields I've used for running my own HYPs, I've often noticed a strong tendency for them all to be moving in the same direction, whereas the actual outcomes are of course that there are big variations between companies about what actually happens, generally with most companies up by about 0-10%, some by >10%, some down by >25% but very few down by 0-25% (excluding currency effects - i.e. with dividend changes calculated in the declaration currency).

Another feature of forecasts is that they're for the next as-yet-not-fully-declared company financial year, which is typically anything from the year which started about a month before the date of the forecast and ends about 11 months afterwards, to the year that starts about 15 months before the date of the forecast and ends about 3 months before, depending on when company financial years end and how quick they are to get their final results out (typically 1-3 months after the end of the financial year). That leads to seasonal changes to the '1-year forecasts', with a lot of changes happening during 'final results seasons' not because the forecasters have changed their minds, but because what was the '2-year forecast' abruptly becomes the '1-year forecast' as each company's final results are announced (*). As a result of that and the large number of companies (especially large international companies) that use the calendar year as their financial year, late January, February and March often have large forecast dividend changes, not necessarily because the forecasters have significantly altered their views, but because the forecast has shifted to a different year.

What that does say to me is that the general picture your chart is giving of variations in dividend income is a combined picture of the variations in actual dividends, forecaster effects, seasonal effects, DL errors and possibly others I haven't thought of. It would be interesting to see an equivalent chart (or another line on the same chart) for what the declared dividends actually turned out to be, especially if they were made to cover precisely the same dividends as the forecasts at any particular date - that's easier to determine after the fact than what the forecasts were, as dividend histories are readily available and forecast-dividend histories are not (at least so far as I know!).

I do appreciate though that it would probably be quite a bit of work, and not possible at all for dates less than about a year in the past (as "what the dividends turned out to be" will not be known until up to about that long ago for any one company, and a large portfolio will probably contain at least one company that is close to that maximum). So I'm not suggesting that you actually try it right now for the particularly odd-looking region on your chart, nor that you do so in the future unless you think it interesting enough to see just how much of the forecast-dividends/unit variation you see in the chart is due to actual dividends/unit variation and how much is due to the use of forecast dividends rather than actual dividends to be worth the effort... I'm not suggesting that forecasts are often grossly wrong when looked at on a whole-portfolio basis for a reasonably large portfolio, by the way - I've never known them to be more than a few percent out - but given that the forecast-dividends variation visible in your chart is only about 5% either side of about 6.75p/unit, a few percent variation due to using forecast dividends rather than actual dividends could account for quite a lot of it.

Gengulphus

[Gengulphus]

There seems to have been some correlation, at least in my portfolio, between USD declarers and flat native (i.e. USD) dividends, e.g. HSBC on a $0.10/$0.10/$0.10/$0.21 cycle for the last 3 years and similar "stuck" dividends for Shell ($0.47 every quarter for the last 4 years), possibly BP but I can't be bothered to do the research on that one and maybe others I have forgotten. ...

Yes, BP has been "stuck" on $0.10c per quarter for 3.5 years now. Most quickly seen "from the horse's mouth" in the company's PDF of the ADS dividend history, though the figures in it do need dividing by 6 because each ADS represents 6 shares.

Gengulphus

[Gengulphus]

Another feature of forecasts is that they're for the next as-yet-not-fully-declared company financial year, which is typically anything from the year which started about a month before the date of the forecast and ends about 11 months afterwards, to the year that starts about 15 months before the date of the forecast and ends about 3 months before, depending on when company financial years end and how quick they are to get their final results out (typically 1-3 months after the end of the financial year). That leads to seasonal changes to the '1-year forecasts', with a lot of changes happening during 'final results seasons' not because the forecasters have changed their minds, but because what was the '2-year forecast' abruptly becomes the '1-year forecast' as each company's final results are announced (*). As a result of that and the large number of companies (especially large international companies) that use the calendar year as their financial year, late January, February and March often have large forecast dividend changes, not necessarily because the forecasters have significantly altered their views, but because the forecast has shifted to a different year.

To illustrate this, I looked through the archived data I have on GDHYP's past share selections, as they record DL forecasts I took at the time I started the selections concerned. I found a suitable set of them covering roughly the 5 months November 2015 - March 2016, in the 32nd, 33rd, 34/35/36th (done together because of the large amount of cash dumped on the portfolio by the Amlin takeover) and 37th selections.

There are a total of 27 companies whose forecasts are recorded in those four start-of-selection threads. Of those, 25 are recorded in all four threads, the exceptions being Amlin (disappeared due to takeover after the first two) and Legal & General (added to GDHYP in the second and so appeared in the third). In addition, two companies had DL forecasts replaced by forecasts I made myself - BLT in the third because its interim had just been severely cut and the forecast had clearly not caught up with events, and GSK in at least the third and possibly others as well because the company had said it intended to pay 80p/year for the next few years and the forecasts were noticeably different from 80p. And two more (FirstGroup and Tesco) were not paying dividends, making the forecasts very small, the problem with which (for the analysis that follows) is that insignificant differences in forecast dividend amounts become big percentage changes to them - FirstGroup's forecasts were 0.22p->0.45p->0.61p->0.65p, which are +105%, +36%, +7% changes, and Tesco's were 0.46p->0.49p->0.27p->0.29p, which are +7%,-45%, +7% changes. (Tesco only restarted about 4-5 months ago, and FirstGroup still hasn't.)

Restricting attention to the other 21 companies, here are the DL dividend forecasts for the dates of the four threads:


In that table, the %change columns show the percentage changes between the forecast dividend figures for the dated columns to their left and right, together with an indication of what results were released between the dates of those columns: normal font = no results released, italics = interim or quarterly results released, bold italics = final results released.

It's quite noticeable that the final results are mostly associated with quite significant forecast-dividend changes - typically more than 3% in one direction or the other - and everything else is mostly associated with much less significant ones - typically under 2% in one direction or the other, and often under 1%. There are exceptions to both of those observations, but they're not very common. And in particular, the exceptions to the second of them are AMEC Foster Wheeler, which was a supplier to miners, a sector in particular turmoil at the time, and Lloyds and RSA, which were both in the fairly early stages of rebuilding their dividends, and also had both released their final results on 25/02/2016, just two days before the 27/02/2016 date of their third forecast. That probably means that the analyst forecast changes that occurred because of the results hadn't (or hadn't fully) worked their way through to the consensus forecasts by 27/02/2016, instead working their way through in the next few days or weeks, and so actually resulting in changes between the 3rd and 4th forecasts.

The big changes associated with final results are doubtless caused by a combination of forecasts shifting from ones for 2015 to 2016 and analysts updating their forecasts after the results. In particular, it's pretty common for analysts to come up a 2-year forecast that is in the region of 5% above their 1-year forecast, so the fact that many of the percentage changes associated with final results suggests that they're probably for companies whose results were more-or-less as expected, so that analysts had little reason to revise their previous forecasts and the main factor in the percentage change was the year shift. But Lloyds in particular indicates that at least sometimes, there were significant analyst forecast revisions more than a couple of days after results (and so not directly caused by the year shift).

Gengulphus