Donate to Remove ads

Got a credit card? use our Credit Card & Finance Calculators

Thanks to tjh290633,csearle,jfgw,bionichamster,Shelford, for Donating to support the site

Article in Telegraph on ISA Millionaires

Investment discussion for beginners. Why you should invest your money, get help getting started
tjh290633
Lemon Quarter
Posts: 3661
Joined: November 4th, 2016, 11:20 am
Has thanked: 245 times
Been thanked: 1238 times

Re: Article in Telegraph on ISA Millionaires

#208128

Postby tjh290633 » March 16th, 2019, 11:34 pm

tjh290633 wrote:
OZYU wrote:One way to see if you are chasing yield too much is to, over a sensibly longish rolling period, see if your income unit growth is persistently lagging your divi per unit growth. If it is lower your yield by gradually altering your portfolio growth/income mix. We keep an eye on such things and measure all performamnce net of RPI.


I will have a look later at the changes over, say 5 or 10 year periods, to see if a a pattern emerges. You will note, of course, that unit value often lags dividend/unit, especially in the 2007-15 period. During that period I was obliged to make adjustments to the portfolio after the event, which was to comparatively lower yield shares from what had previously been higher yield shares, but which failed to live up to their name.

TJH

I have now had a chance to look at the changes over 5 year periods:

.            .          .                   5 Yr Change                        
Year to Unit Value Div/Unit Unit Value Div/Unit Difference
21-Apr-87 1.00 0.00
05-Apr-88 0.92 2.87
05-Apr-89 1.19 2.75
05-Apr-90 1.24 4.33
05-Apr-91 1.42 5.75 41.65%
05-Apr-92 1.38 7.97 50.86% 177.71% 349.42%
05-Apr-93 1.60 7.33 34.49% 166.74% 483.48%
05-Apr-94 1.81 6.65 45.97% 53.33% 116.00%
05-Apr-95 1.75 7.93 23.87% 37.83% 158.51%
05-Apr-96 2.07 7.81 49.50% -2.00% -4.05%
05-Apr-97 2.29 8.90 43.19% 21.44% 49.64%
05-Apr-98 3.48 10.52 92.79% 58.35% 62.89%
05-Apr-99 3.62 8.91 106.45% 12.33% 11.58%
05-Apr-00 3.51 11.96 69.84% 53.10% 76.02%
05-Apr-01 3.48 13.15 51.96% 47.73% 91.86%
05-Apr-02 3.57 13.82 2.63% 31.27% 1187.16%
05-Apr-03 2.45 12.95 -32.31% 45.46% -140.72%
05-Apr-04 3.14 14.37 -10.50% 20.10% -191.39%
05-Apr-05 3.72 14.02 6.80% 6.59% 96.92%
05-Apr-06 4.62 18.70 29.26% 35.33% 120.77%
05-Apr-07 5.27 20.84 114.95% 60.84% 52.93%
05-Apr-08 4.44 26.09 41.33% 81.62% 197.49%
05-Apr-09 2.45 22.76 -34.04% 62.40% -183.35% *
05-Apr-10 3.94 11.91 -14.61% -36.31% 248.54% *
05-Apr-11 4.61 16.71 -12.53% -19.78% 157.85% *
05-Apr-12 4.74 19.09 6.74% -26.83% -397.88%
05-Apr-13 5.27 22.91 115.01% 0.63% 0.55%
05-Apr-14 5.61 24.19 42.25% 103.16% 244.20%
05-Apr-15 6.21 26.23 34.63% 56.95% 164.48%
05-Apr-16 5.92 23.81 25.05% 24.73% 98.73%
05-Apr-17 6.62 26.21 25.53% 14.42% 56.47%
05-Apr-18 6.12 33.19 9.07% 37.17% 409.74%
16-Mar-19 6.25 31.57 0.69% 20.35% 2945.60%

Figures less than 100% in the final column indicate that the income lagged the unit price in that 5 year period. In a few instances marked (*) there are spurious figures caused by a negative change in the unit price over that period, or by both changes being negative. A rampant bull market will lead to the growth in unit price outpacing the growth in dividend, whereas a bear market will have the opposite effect.

As I said previously, dividend changes lag price changes because they refer to previous events in the main. Hence the 2008 crisis was not reflected in dividends immediately.

TJH

OZYU
2 Lemon pips
Posts: 199
Joined: December 31st, 2016, 3:52 pm
Has thanked: 43 times
Been thanked: 138 times

Re: Article in Telegraph on ISA Millionaires

#208161

Postby OZYU » March 17th, 2019, 10:36 am

Hi TJ,

Our aims are different, I don't and never will need the income. So any advice I might give is directed at the generally younger pot builder who needs to be mindful of building his pot, hence tr. As you know we, my wife and I, run five different portfolios, and only seek some higher yield in the ISAs, for obvious reasons.


I am always astonished at the large variations on your data too, much much higher than mine. We have also disagreed off board on some aspects of unitisation anyway.

So comparing my data with yours would be silly. Your average yield over the period is more than 1% higher than mine( in my ISA), but the my overall Inc unit and divi per unit CAGRs are much closer to each over. Your are over the years were there is data in both columns, 6.4% and 8.1% respectively for you overall( mine, in the ISA, are 8.93% and 7.95% at Fridays close over the same period)). That in itself indicates that you are indeed chasing yield, but not excessively bearing in mind your aim, which it to take plenty of income.

Now for the analysis. I do it differently. Calculating the ten year rolling CAGRs of both on your data, and the ratios, produces this: Sorry no time to present better, as we are getting ready to go out to some friends for lunch some way away.


10 year to try to smooth out things better.

10 year CAGRs on your data
Unit D, per Unit, (1+U)/(1+D)-1 in the last column.
14.2% 14% 0.3%
11.8% 12% -0.6%
11.0% 11% 0.2%
9.4% 9% 0.7%
10.0% 6% 4.1%
4.4% 6% -1.4%
5.7% 8% -2.2%
7.8% 6% 1.9%
8.4% 9% -0.7%
8.7% 9% -0.2%
2.5% 9% -6.4%
-3.8% 10% -12.4%
1.2% 0% 1.2%
2.9% 2% 0.4%
2.9% 3% -0.4%
8.0% 6% 2.0%
6.0% 5% 0.6%
5.3% 6% -1.1%
2.5% 2% 0.1%
2.3% 2% 0.0%
3.3% 2% 0.8%
9.9% 3% 6.3%
As you might see, rogue data still present because of your deep financial crisis deep income glitch, but the pattern is ok ish.


My conclusion, your are running a fairly balanced portfolio, no surprise to anybody here. And like us, the recent decades have not been a patch on the early ones.


Hope this helps


Ozyu

Gengulphus
Lemon Quarter
Posts: 3008
Joined: November 4th, 2016, 1:17 am
Been thanked: 1554 times

Re: Article in Telegraph on ISA Millionaires

#208199

Postby Gengulphus » March 17th, 2019, 2:10 pm

tjh290633 wrote:... In a few instances marked (*) there are spurious figures caused by a negative change in the unit price over that period, or by both changes being negative. ...

Shouldn't the rows for 05-Apr-03 and 05-Apr-04 have been asterisked as well? (If not, I'm failing to understand what you're calculating.)

But I would suggest that figures being spurious in that way is a reason to mistrust that what you're calculating is the right thing to calculate. And on thinking about it a bit more, the figures you would get if the unit price changes were lagging the div/unit changes are very similar to those you would get if the div/unit changes were lagging the unit price changes. E.g. consider the following 'toy' examples, in which the basic figures rise by 100% each year except for one, when they rise by 50%:

Unit price   Div/unit   'Difference'  |  Unit price   Div/unit   'Difference'
20 1 | 20 1
40 (+100%) 2 (+100%) 100% | 40 (+100%) 2 (+100%) 100%
60 (+50%) 4 (+100%) 200% | 80 (+100%) 3 (+50%) 50%
120 (+100%) 6 (+50%) 50% | 120 ( +50%) 6 (+100%) 200%
240 (+100%) 12 (+100%) 100% | 240 (+100%) 12 (+100%) 100%

The first set of columns has the div/unit change lagging the unit price change; the second has the unit price change lagging the div/unit change - but they both produce 'difference' figures of one 50%, one 200% and two 100%s. So rather deeper analysis of those 'difference' figures is needed to tell which is lagging the other.

But actually, I'd suggest using correlation coefficients. If for example unit price changes were lagging div/unit changes by 2 years, you would expect the unit price column to be well-correlated with the div/unit column offset downwards in the table by 2 years. I've done that for offsets in both directions of up to 5 years - it's a very simple process in a spreadsheet, just involving applying the CORREL function to ranges of your Unit Value and Div/Unit columns which are offset from each other. The resulting correlation coefficients are:

Unit price lagging div/unit by 5 years:  0.8096
Unit price lagging div/unit by 4 years: 0.8221
Unit price lagging div/unit by 3 years: 0.8047
Unit price lagging div/unit by 2 years: 0.8247
Unit price lagging div/unit by 1 year: 0.8358
No lag in either direction: 0.9008
Div/unit lagging unit price by 1 year: 0.9609
Div/unit lagging unit price by 2 years: 0.9283
Div/unit lagging unit price by 3 years: 0.8563
Div/unit lagging unit price by 4 years: 0.8063
Div/unit lagging unit price by 5 years: 0.7864

The biggest of those correlation coefficients is the one for div/unit lagging unit price by a year, so that's the hypothesis best supported by your data. At least to the nearest year - the facts that a 2-year lag has a somewhat higher correlation coefficient than no lag, and that a 3-year lag a somewhat higher correlation coefficient than a 1-year lag in the opposite direction, suggest that a closer examination using years starting in different months might say that a lag of a few months over a year was best supported. (But I'm not suggesting that that would be a worthwhile use of one's time!)

Gengulphus

tjh290633
Lemon Quarter
Posts: 3661
Joined: November 4th, 2016, 11:20 am
Has thanked: 245 times
Been thanked: 1238 times

Re: Article in Telegraph on ISA Millionaires

#208243

Postby tjh290633 » March 17th, 2019, 6:14 pm

Gengulphus wrote:
tjh290633 wrote:... In a few instances marked (*) there are spurious figures caused by a negative change in the unit price over that period, or by both changes being negative. ...

Shouldn't the rows for 05-Apr-03 and 05-Apr-04 have been asterisked as well? (If not, I'm failing to understand what you're calculating.)

But I would suggest that figures being spurious in that way is a reason to mistrust that what you're calculating is the right thing to calculate. And on thinking about it a bit more, the figures you would get if the unit price changes were lagging the div/unit changes are very similar to those you would get if the div/unit changes were lagging the unit price changes. E.g. consider the following 'toy' examples, in which the basic figures rise by 100% each year except for one, when they rise by 50%:

Yes, you are right. I thought up a better was to do the calculation while lying in bed last night. I am being called for food, so another time.

TJH

tjh290633
Lemon Quarter
Posts: 3661
Joined: November 4th, 2016, 11:20 am
Has thanked: 245 times
Been thanked: 1238 times

Re: Article in Telegraph on ISA Millionaires

#208471

Postby tjh290633 » March 18th, 2019, 6:24 pm

I have now got round to revising my 5-year table. What came to me was that, if I subtracted the unit price change from the dividend per unit change, then I got a correct answer.


Now negative figures in the final column indicate where the dividend has lagged the unit price overf the 5 years in question. 14 altogether, over 27 periods.

TJH

Gengulphus
Lemon Quarter
Posts: 3008
Joined: November 4th, 2016, 1:17 am
Been thanked: 1554 times

Re: Article in Telegraph on ISA Millionaires

#208501

Postby Gengulphus » March 18th, 2019, 8:43 pm

tjh290633 wrote:What came to me was that, if I subtracted the unit price change from the dividend per unit change, then I got a correct answer.

Ah! I think you and I have been talking somewhat at cross-purposes, because there are two different possible meanings of "lag" here. You're talking about dividends lagging in the sense that the percentage increase in the dividend is less than the percentage increase in the unit price. Or almost equivalently, the yield has dropped between the beginning and end of the period, since:

(ending yield) < (starting yield) is equivalent to:

(div-per-unit at end) / (unit price at end) < (div-per-unit at start) / (unit price at start) by the definition of yield, which is almost equivalent to:

(div-per-unit at end) / (div-per-unit at start) < (unit price at end) / (unit price at start) by multiplying through by (unit price at start) / (div-per-unit at start), which is positive provided (div-per-unit at start) is non-zero, which is equivalent to:

1 + (percentage increase in div-per-unit)/100 < 1 + (percentage increase in unit price)/100 by the definition of percentage increases, which is equivalent to:

(percentage increase in div-per-unit) < (percentage increase in unit price) by subtracting 1 from both sides and then multiplying both sides by 100.

And the two are also basically equivalent in the special case that (div-per-unit at start) = 0, since in that case the starting yield is zero and it's impossible for the ending yield to be negative, but the percentage increase in div-per-unit is either infinite or indeterminate, and in either case it's impossible for it to be less than the percentage increase in unit price.

But I was taking "lag" to mean "behind in time" - i.e. basically, that especially large or small increases in one tend to happen a certain amount of time after particularly large or small increases in the other. The two concepts do interact, but not in a very straightforward way: if e.g. the div-per-unit increases are tending to happen a year later than the unit price increases (i.e. 'lagging' in my sense), then a particularly large unit price increase in the last year of a 5-year period will tend to boost the percentage increase in the unit price but not the percentage increase in div-per-unit (so to show up as 'lagging' in your sense), but a particularly small unit price increase in the same year will do the opposite (so to show up as the opposite of 'lagging' in your sense). And similarly, those same changes in the year previous to the first year of a 5-year period will tend to show up as the opposite of 'lagging' in your sense if particularly large and as 'lagging' in your sense if particularly small.

Not saying that either concept is better - which is better depends on what you want to know. Just that they are quite different concepts from each other, and so mistaking one for the other is a recipe for confusion!

Gengulphus

LooseCannon101
2 Lemon pips
Posts: 123
Joined: November 5th, 2016, 2:12 pm
Has thanked: 168 times
Been thanked: 68 times

Re: Article in Telegraph on ISA Millionaires

#208966

Postby LooseCannon101 » March 20th, 2019, 9:29 pm

I completely agree with the Telegraph article.

Most people could achieve a return of 8% per annum over 20+ years with a low-cost world equity index tracker or equivalent investment trust(s).

Unfortunately, the average retail investor wastes time and money chasing the latest fad and getting burnt in the process. Stockbrokers encourage this foolish behaviour because it increases their profits.

Sitting on one's hands and re-investing dividends is extremely boring and requires nerves of steel at times, but over the long run will deliver a substantial increase in wealth.


Return to “How Do I Invest”

Who is online

Users browsing this forum: No registered users and 1 guest