QE and index funds have deformed the stock market as predicted by Goodhart’s Law.

[OhNoNotimAgain]

No, its compound interest. For that to work you need income. I have yet to see a credible study that shows capital growth is equally important to dividends.

Capital growth has been a jolly sight more important than dividends for Berkshire Hathaway, which pays no dividends. Some of the tech giants did not pay dividends for a long time either. Share buy backs are clearly every bit as important as dividends. Total return is what matters.

So why do you think BH doesn't pay dividends but invests in companies that do?

[Bubblesofearth]

The British market is irrelevant here. It is only about 4.1% of the global market. You will not get much diversification there. If you hold international shares directly, you face high exchange rate costs. You also have the problem of reclaiming withholding tax in umpteen languages. Your diversification will still be poor unless you hold thousands of shares. Interactive Brokers is relatively cheap, but they do not do ISAs or SIPPs. Anyway, what is the point? if the professionals cannot beat the market, except by chance, what chance do you have?

The UK market may be small but the point remains that if you are investing only in that market then, for the reasons outlined in my previous post, an equal weight on purchase long term buy and hold strategy should beat the relevant index (FTSE all share)

Yes, ideally for maximum diversification an investor would be advised to buy shares internationally. And, yes, there are problems with such an approach. One middle ground is to buy a range of different country, or regional, trackers in equal amounts rather than a global tracker. I did try a bit of back-testing of this approach and found significant outperformance vs the Global tracker for the period covered by Dimson's book 'Triumph of the optimists' from which I took the data.

As regards professionals not beating the market, part of that is down to higher charges and part down to it being very difficult for most of them to adopt the kind of LTBH approach that PI's can use. It would be interesting to see the approach taken by wealth managers offering bespoke services. I don't have much data there.

BoE

[Bubblesofearth]

No, its compound interest. For that to work you need income. I have yet to see a credible study that shows capital growth is equally important to dividends.

This is a pretty comprehensive review of asset returns;

https://economics.harvard.edu/files/eco ... s28533.pdf

From which on p.39;

The importance of dividends and rents is partly a matter of convention. Appendix N and Appendix Table A.20 computes the equivalent decomposition for nominal returns, and finds that the capital gain versus dividend/rental income split is then closer to roughly 50/50.


Be careful with some of the earlier tables which show real cap growth, i.e. after inflation. It may be that other studies also show this, whilst showing nominal dividend returns, which will of course lead to the erroneous conclusion that dividends are more important than cap growth in driving total return.

BoE

[GoSeigen]

Be careful with some of the earlier tables which show real cap growth, i.e. after inflation.

That's exactly the trap he fell into the last time I saw him try to make this point.

GS

[OhNoNotimAgain]

No, its compound interest. For that to work you need income. I have yet to see a credible study that shows capital growth is equally important to dividends.

This is a pretty comprehensive review of asset returns;

https://economics.harvard.edu/files/eco ... s28533.pdf

From which on p.39;

The importance of dividends and rents is partly a matter of convention. Appendix N and Appendix Table A.20 computes the equivalent decomposition for nominal returns, and finds that the capital gain versus dividend/rental income split is then closer to roughly 50/50.


Be careful with some of the earlier tables which show real cap growth, i.e. after inflation. It may be that other studies also show this, whilst showing nominal dividend returns, which will of course lead to the erroneous conclusion that dividends are more important than cap growth in driving total return.

BoE

Nominal returns are not much use to anyone.

On page 39 the paper says:

Table VIII decomposes equity and housing returns into capital gains and dividends or rents, for
the full cross-country sample and the period after 1950. Over the full sample, most of the real return
is attributable to the yield. Dividends account for roughly 60% of real equity returns, and rents for
roughly 80% of real housing returns. In terms of geometric means (Table VIII, row 3), almost all of
both equity and housing returns are attributable to, respectively, dividend and rental income

[Bubblesofearth]

Nominal returns are not much use to anyone.

On page 39 the paper says:

Table VIII decomposes equity and housing returns into capital gains and dividends or rents, for
the full cross-country sample and the period after 1950. Over the full sample, most of the real return
is attributable to the yield. Dividends account for roughly 60% of real equity returns, and rents for
roughly 80% of real housing returns. In terms of geometric means (Table VIII, row 3), almost all of
both equity and housing returns are attributable to, respectively, dividend and rental income

The word 'attributable' is key here - it is how the returns have been calculated to allow for inflation. I can only advise you once again to refer to Appendix N.

Alternatively think about it in terms of earnings. Does it seem reasonable to you that companies that retain earnings for growth are going to do much worse than those that pay out earnings as dividends? If dividends are responsible for almost all total return then surely all companies should pay out their earnings as dividends. If not, why not?

BoE

[OhNoNotimAgain]

Nominal returns are not much use to anyone.

On page 39 the paper says:

Table VIII decomposes equity and housing returns into capital gains and dividends or rents, for
the full cross-country sample and the period after 1950. Over the full sample, most of the real return
is attributable to the yield. Dividends account for roughly 60% of real equity returns, and rents for
roughly 80% of real housing returns. In terms of geometric means (Table VIII, row 3), almost all of
both equity and housing returns are attributable to, respectively, dividend and rental income

The word 'attributable' is key here - it is how the returns have been calculated to allow for inflation. I can only advise you once again to refer to Appendix N.

Alternatively think about it in terms of earnings. Does it seem reasonable to you that companies that retain earnings for growth are going to do much worse than those that pay out earnings as dividends? If dividends are responsible for almost all total return then surely all companies should pay out their earnings as dividends. If not, why not?

BoE

It's an asset class story. Dividends paid out by one company are not necessarily reinvested back into the same company. They are reinvested across the whole of the market.

[dealtn]

This has been exposed as a fallacy many times before on these boards. For the market as a whole return from dividends and cap growth are approximately equal. The trap not to fall into is to attribute cap growth on reinvested dividends to dividends. It's cap growth. This is more obvious for growth on retained earnings but comes to the same thing.

The danger inherent in thinking that dividends matter more than growth is that it can lead you down a path of chasing exclusively high dividend payers. The risk of this approach is lack of adequate diversification, both local and international.

BoE

No, its compound interest. For that to work you need income. I have yet to see a credible study that shows capital growth is equally important to dividends.

Bear in mind that the last two decades, I repeat, two decades, are an aberration in capital markets because of super low interest rates and QE. The proof of that is UK gilts have outperformed UK equities over that period.

No its not. Its not even interest!

How different do you think Total Returns would be if profits weren't paid out but retained in the business and that compounding happened internally? Is the maths somehow different? If anything it will be even better as the frictional costs of trading and taxation are less.

[hiriskpaul]

Not this dividends nonsense again. I remember dealing with that one multiple times before.

From memory I think the S&P 500 derived around 25% of its return from dividends, the rest from capital growth. The higher yielding FTSE all share about 50%. Completely agree with the comments that silly games can be played to make dividends look more important, such as incorrectly attributing the capital growth on shares bought with dividends to the dividends and to only apply inflation to capital growth.

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Compound interest works with interest payments but does not work with dividend payments because of these share price drops with each dividend payment. Each dividend reinvestment does not happen at the precise (share price - dividend per share) of course. Sometimes the reinvestment will take place at a higher price, sometimes lower, but that is what it averages out at.

The whole dividend compound interest thing is a nonsense at another level. If a company pays out a dividend, it is impossible for the market as a whole to reinvest that dividend because it requires a source of shares to be bought. Clearly everyone holding shares in a company cannot reinvest as some shareholders must sell to match with the reinvestors.

[absolutezero]

Not this dividends nonsense again. I remember dealing with that one multiple times before.
...
To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Factor in dealing costs and spreads and you are worse off than you started!

Compound interest works with interest payments but does not work with dividend payments because of these share price drops with each dividend payment. Each dividend reinvestment does not happen at the precise (share price - dividend per share) of course. Sometimes the reinvestment will take place at a higher price, sometimes lower, but that is what it averages out at.

The whole dividend compound interest thing is a nonsense at another level. If a company pays out a dividend, it is impossible for the market as a whole to reinvest that dividend because it requires a source of shares to be bought. Clearly everyone holding shares in a company cannot reinvest as some shareholders must sell to match with the reinvestors.

Indeed. Dividends are not free money and are not akin to interest at all.

[OhNoNotimAgain]

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Taking your argument to its logical conclusion a share would therefore fall every time a dividend was paid until it reached zero.

[GeoffF100]

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Taking your argument to its logical conclusion a share would therefore fall every time a dividend was paid until it reached zero.

hiriskpaul is absolutely correct as I am sure you know well. If a company were to pay out its entire market capitalisation as a dividend without earning any new money, it would indeed be worthless.

[OhNoNotimAgain]

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Taking your argument to its logical conclusion a share would therefore fall every time a dividend was paid until it reached zero.

hiriskpaul is absolutely correct as I am sure you know well. If a company were to pay out its entire market capitalisation as a dividend without earning any new money, it would indeed be worthless.

Terry, please help and explain operating cash flow to this board.

I

[GoSeigen]

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Taking your argument to its logical conclusion a share would therefore fall every time a dividend was paid until it reached zero.

hiriskpaul is absolutely correct as I am sure you know well. If a company were to pay out its entire market capitalisation as a dividend without earning any new money, it would indeed be worthless.

I think he was making an absurd point to match the absurdity of ONNA's recurrent mathematical fakery. Well I laughed at it anyway...

GS

[hiriskpaul]

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Taking your argument to its logical conclusion a share would therefore fall every time a dividend was paid until it reached zero.

??

Do I really have to explain how companies generate dividends?

[scrumpyjack]

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Taking your argument to its logical conclusion a share would therefore fall every time a dividend was paid until it reached zero.

hiriskpaul is absolutely correct as I am sure you know well. If a company were to pay out its entire market capitalisation as a dividend without earning any new money, it would indeed be worthless.

Well, no, it depends. For example Caledonia Investment Trust has a share price of 3675p but a net asset value of about 4900p. If it paid out a dividend of 3675p per share, it would not be worthless. There would still be net assets per share of 1225p!

[hiriskpaul]

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Taking your argument to its logical conclusion a share would therefore fall every time a dividend was paid until it reached zero.

hiriskpaul is absolutely correct as I am sure you know well. If a company were to pay out its entire market capitalisation as a dividend without earning any new money, it would indeed be worthless.

Well, no, it depends. For example Caledonia Investment Trust has a share price of 3675p but a net asset value of about 4900p. If it paid out a dividend of 3675p per share, it would not be worthless. There would still be net assets per share of 1225p!

The moment they announced a 3675p distribution the share price would leap because of the certainty of that near cash flow. When it went XD though, the price would come down 3675p.

[OhNoNotimAgain]

OK guys, you win.
I give up.
Sayonara

[GoSeigen]

To take another view I could argue that dividends contribute absolutely nothing to long term returns. The reason is that on average with each dividend you get an equal and opposite drop in the share price. Reinvest that dividend at the reduced share price and your total holding value is identical to that had the dividend never been paid. In other words, you are back where you started before the dividend was paid. More shares, but at a lower price per share. Before anyone objects, I suggest they do the calculation. If you don't get back to where you started you have done it wrong!

Taking your argument to its logical conclusion a share would therefore fall every time a dividend was paid until it reached zero.

hiriskpaul is absolutely correct as I am sure you know well. If a company were to pay out its entire market capitalisation as a dividend without earning any new money, it would indeed be worthless.

I think he was making an absurd point to match the absurdity of ONNA's recurrent mathematical fakery. Well I laughed at it anyway...

It's worth saying that I have never taken issue with the notion that dividends could account for more than 50% of equity investment returns. Perhaps they do, either through conducive conditions over the investing period or some underlying fact that ensures such an outcome.

My quarrel with OhNoNotimAgain has only been the bizarre maths he used to try to demonstrate the point -- a case of (possibly) right answer but nonsense working.


GS

[tjh290633]

Terry, please help and explain operating cash flow to this board.

The thing that seems to be missed is that both dividends and capital growth contribute to the total return of a share. Here is a little example based on IGG, showing 3 buys of a single share and the dividends which would have accrued:

IGG                                                      IRR    
Date         Event    Price £   Dividend p   No Shares     1.01%
29/12/2020   Bought   9.0208                     1.00    -9.0208
25/02/2021   H1       8.5550        12.96                 0.1296
14/06/2021   Bought   8.7051                     2.00    -8.7051
21/10/2021   FY       8.5750        30.24                 0.6048
14/01/2022   Bought   8.0450                     3.00    -8.0450
04/03/2022   H1       8.5000        12.96                 0.3888
20/10/2022   FY       7.5200        31.24                 0.9372
27/09/2022   Sold     8.0100                     0.00    24.0300

As you can see, quite a lot of capital loss, but some dividends accrued. The IRR is just positive, because the cash flow in the right hand column is just positive. In this case the dividends have added to the cash flow to make its total just positive to the extent of 32p.

TJH