Tilting for yield and the bath analogy

[mc2fool]

Dimson et al's main body of work isn't specifically about 2) but it is true in this specific paper of 2011 they explore it and agree with it. CRUCIALLY though this long term outperformance is driven by 2 factors 1) Higher Risk and 2) Buying Value (and avoiding premium priced alternatives, and holding until that value is outed then repeated.

The paper doesn't really agree with the "higher risk" aspect. They look at standard deviations, betas and Sharpe ratios and conclude: "It is therefore hard to explain the superior performance of yield-tilt strategies in terms of risk, at least as conventionally defined. Indeed, when growth and value stocks are defined based on dividend yield, it is the value stocks that have the lower volatility and beta." See pages 20-21.

Perhaps related is the oft quoted matter that "defensives", which are often dull lower risk tortoise like shares, tend to outperform over the long term. (Maybe another mispricing?) https://www.investorschronicle.co.uk/comment/2019/10/17/why-defensives-outperform/

[dealtn]

Dimson et al's main body of work isn't specifically about 2) but it is true in this specific paper of 2011 they explore it and agree with it. CRUCIALLY though this long term outperformance is driven by 2 factors 1) Higher Risk and 2) Buying Value (and avoiding premium priced alternatives, and holding until that value is outed then repeated.

The paper doesn't really agree with the "higher risk" aspect. They look at standard deviations, betas and Sharpe ratios and conclude: "It is therefore hard to explain the superior performance of yield-tilt strategies in terms of risk, at least as conventionally defined. Indeed, when growth and value stocks are defined based on dividend yield, it is the value stocks that have the lower volatility and beta." See pages 20-21.

Perhaps related is the oft quoted matter that "defensives", which are often dull lower risk tortoise like shares, tend to outperform over the long term. (Maybe another mispricing?) https://www.investorschronicle.co.uk/comment/2019/10/17/why-defensives-outperform/

I agree 1) is low, and actually hard to quantify.

One challenge, and one I don't recall ever seeing empiracally resolved, was the fact that a number of low (or no) yield stocks were from ex-high yield ones that suspended payouts. The data found it hard to make a comparison between (a lazily labelled by me) "growth stocks" and "high dividend stocks" when the former's population got "polluted" by some of the worst performing of the latter that suspended dividends.

Its genuinely hard to do, as those stocks will feature among them many that go bust, but also many that spectacularly recover. What is known is that some of the "riskiness" of those companies get transferred from the latter basket to the former basket at the times dividends get suspended, and the increased volatility of return of them gets transferred to the former too.

[Bubblesofearth]

However, if you want to try asking them I believe DM&S are contactable, at least they were. I recall a time on TMF when it had been picked up, either from this paper or some other, that DM&S had made a statement along the lines of the infamous "most stockmarket returns come from dividends" (not an exact quote! And I think there was a "real" in there, IIRC) and some Fool emailed them to challenge that and one of them (can't remember which) replied and the Fool posted their reply (which basically doubled down) on TMF.

(OT: I do wish the Wayback Machine would implement text search through their archive!)

Hi mc2fool

I was that Fool (then posting as 'lookingforclues). Following (yet another) debate on TMF about the relative contribution of dividends and cap growth to total return I wanted DM&S to confirm that dividends and capital growth were responsible for a roughly 50-50 split in their contribution for the UK market. Unfortunately I didn't word my question precisely enough and the response was to give over-riding importance to reinvested dividends. They demonstrated this by showing the return without reinvested dividends compared to that achieved with them. Clearly this is huge but it completely missed the point I was trying to get at! To use dealt's bath analogy it's like comparing the volume of water refilling from the bucket vs throwing the bucket water away.

I did try to go back to DM&S with what I thought was a more precise wording of the question but they'd obviously had enough of me cos I never got a further response.

I tend to agree with what dealtn has written, that much of the outperformance of a value strategy (value here defined by DM&S as >50% average market yield) is down to repricing of cap values. You can actually see this in operation for HYP1 where tobacco stocks drove much of the return. These were out-of-favour high yielders back in 2000.

One thing maybe worth mentioning is the degree of skew in market returns, the fact that the bulk of returns have tended to come from a minority of stocks. This has interesting implications for strategy, level of diversification and maybe even impacts on why there has historically been a yield tilt. It's a feature of markets that is often overlooked when considering averages.

BoE

[Lootman]

One thing maybe worth mentioning is the degree of skew in market returns, the fact that the bulk of returns have tended to come from a minority of stocks. This has interesting implications for strategy, level of diversification and maybe even impacts on why there has historically been a yield tilt. It's a feature of markets that is often overlooked when considering averages.

Indeed. The US market is relatively broad but even so it is dominated by a relatively small number of shares. If they go up, the index is up, and vice versa.

So for example the top 6 shares by market cap are Apple, MicroSoft, Amazon, Google, Tesla and Berkshire Hathaway. Together they are about 20% of the S&P 500. Moreover they are mostly more recent issues meaning that their market cap is their gain in market cap in recent decades. If you missed out on those 6 you would have been very lucky to have out-performed the US index, and therefore in turn the global index.

Of those 6, 4 have never paid a dividend and 2 pay out a smallish dividend (or more accurately are on a low yield).

So if you took a Ohno/Munro approach to investing and gave a zero weight to these shares since inception, effectively regarding them as worthless, then would you not be doomed to under-perform, at least if you use a cap-weighted benchmark? And even more so when you consider that many of the companies in the top 30 or 40 are also non dividend payers e.g. Netflix, Facebook, Adobe, Nvidia, SalesForce and so on.

[mc2fool]

However, if you want to try asking them I believe DM&S are contactable, at least they were. I recall a time on TMF when it had been picked up, either from this paper or some other, that DM&S had made a statement along the lines of the infamous "most stockmarket returns come from dividends" (not an exact quote! And I think there was a "real" in there, IIRC) and some Fool emailed them to challenge that and one of them (can't remember which) replied and the Fool posted their reply (which basically doubled down) on TMF.

(OT: I do wish the Wayback Machine would implement text search through their archive!)

Hi mc2fool

I was that Fool (then posting as 'lookingforclues).

Aha!!!! LFC!!!! I was hoping whoever it was would still be around and spot my comment and pipe up, if nothing else to prove that I wasn't making it up or going gaga! So thanks for that! (And for enlightening me that BoE = LFC ... I hadn't realised!)

[Itsallaguess]

I'm struggling to see how the bath analogy can be stated by you to not consider 'value', when it's being specifically provided as an analogy to debunk a statement that clearly does.

Can you help bridge that gap in my understanding?

Honestly the bath analogy, which is mine, not Dimson et al's, is really to focus the mind on what drives "volume" of water. Which is the taps (water in) and how much water escapes.

Its simple point is that many focus on the dividend streams, which are the incomes "in" to the shareholders, when actually the dividends are the "money" out from the company's perspective. Of course from a shareholder perspective it is Total Return that matters and obviously the dividends received are a key component of that Total Return, but they are just a transformation from Capital into Income - a net zero (or worse due to frictional costs) movement of shareholder worth from bath to bucket.

Arguments made, by Rob and others, revolve around 1) the Dividends being the driver of shareholder returns (they aren't its company earnings that are), or 2) reinvestment of those dividends is what drives outperformance (which isn't true as they would underperform at the shareholder level the exact same company with no dividends, complete retained earnings, and no recycling of dividends to reinvest.

What is usually behind the argument, once teased away is 1) A portfolio with reinvested dividends exceeds one where dividends aren't reinvested. So obviously true its barely worth discussing or 2) High Dividend strategies outperform other strategies on a long term Total Return basis (as a generalisation).

Dimson et al's main body of work isn't specifically about 2) but it is true in this specific paper of 2011 they explore it and agree with it. CRUCIALLY though this long term outperformance is driven by 2 factors 1) Higher Risk and 2) Buying Value (and avoiding premium priced alternatives, and holding until that value is outed then repeated.

The bath analogy as stated has nothing to say about this, and as I said I think it is tortuous to extend it beyond its original intention.

So if we purely concentrate on the bath analogy for now then, am I right in thinking that it's really only trying to describe two 'twin' businesses, where one of those business-twins reinvests it's own capital 'back into itself', and the other business-twin chooses to pay out more capital in the form of dividends, and where the subsequent re-investment of those dividends back into the same source business must then generate frictional costs that, ultimately, can only ever be at the detriment of the investor owning shares in that second dividend-paying business, compared to an investor owning shares in the first business?

Cheers,

Itsallaguess

[EthicsGradient]

If the bath analogy is for comparing 2 twin businesses, then we can't really apply the real-world data for the same problem, because that's not for "twinned" businesses, but for whole stock markets made up of differing companies.

In reality, the advantage or disadvantage of distributing dividends over reinvesting profit in a company will depend on the kind of company it is, and its business. If you want to go back to a bath analogy, what if the bath is already nearly full? If there's little room for growth in the sector, and the company cannot take much share from competitors, then you may as well catch some water in a bucket through the plughole, rather than seeing the bath overflow (or maybe "the flow through the taps backs up and slows down"). Or you have to build a bigger bath, which takes effort.

In a small but growing company (either by being in a brand new sector, or a new competitor with an advantage over the established ones), there's room for growth just by keeping the plug in.

But at some point, an investor wants to be able to get out more cash than they put in. With a no-dividend-but-growing company that's fine, as long as the market in it is still liquid (as it should be, in a bath ...). But what if there's a risk of the company's products becoming outdated? Reinvesting in a company builds up the company infrastructure, which is great if its products still sell, but having shares in the world's largest buggy-whip manufacturer is little consolation if you never received a dividend before those damn horseless carriages came along. Dividends are at least cash, rather than a specialised system that may not be efficiently convertible to making something else.

[dealtn]

I'm struggling to see how the bath analogy can be stated by you to not consider 'value', when it's being specifically provided as an analogy to debunk a statement that clearly does.

Can you help bridge that gap in my understanding?

Honestly the bath analogy, which is mine, not Dimson et al's, is really to focus the mind on what drives "volume" of water. Which is the taps (water in) and how much water escapes.

Its simple point is that many focus on the dividend streams, which are the incomes "in" to the shareholders, when actually the dividends are the "money" out from the company's perspective. Of course from a shareholder perspective it is Total Return that matters and obviously the dividends received are a key component of that Total Return, but they are just a transformation from Capital into Income - a net zero (or worse due to frictional costs) movement of shareholder worth from bath to bucket.

Arguments made, by Rob and others, revolve around 1) the Dividends being the driver of shareholder returns (they aren't its company earnings that are), or 2) reinvestment of those dividends is what drives outperformance (which isn't true as they would underperform at the shareholder level the exact same company with no dividends, complete retained earnings, and no recycling of dividends to reinvest.

What is usually behind the argument, once teased away is 1) A portfolio with reinvested dividends exceeds one where dividends aren't reinvested. So obviously true its barely worth discussing or 2) High Dividend strategies outperform other strategies on a long term Total Return basis (as a generalisation).

Dimson et al's main body of work isn't specifically about 2) but it is true in this specific paper of 2011 they explore it and agree with it. CRUCIALLY though this long term outperformance is driven by 2 factors 1) Higher Risk and 2) Buying Value (and avoiding premium priced alternatives, and holding until that value is outed then repeated.

The bath analogy as stated has nothing to say about this, and as I said I think it is tortuous to extend it beyond its original intention.

So if we purely concentrate on the bath analogy for now then, am I right in thinking that it's really only trying to describe two 'twin' businesses, where one of those business-twins reinvests it's own capital 'back into itself', and the other business-twin chooses to pay out more capital in the form of dividends, and where the subsequent re-investment of those dividends back into the same source business must then generate frictional costs that, ultimately, can only ever be at the detriment of the investor owning shares in that second dividend-paying business, compared to an investor owning shares in the first business?

Cheers,

Itsallaguess

Yes. (Although I dislike absolute words like "only" and "must").

[dealtn]

If the bath analogy is for comparing 2 twin businesses, then we can't really apply the real-world data for the same problem, because that's not for "twinned" businesses, but for whole stock markets made up of differing companies.

In reality, the advantage or disadvantage of distributing dividends over reinvesting profit in a company will depend on the kind of company it is, and its business. If you want to go back to a bath analogy, what if the bath is already nearly full? If there's little room for growth in the sector, and the company cannot take much share from competitors, then you may as well catch some water in a bucket through the plughole, rather than seeing the bath overflow (or maybe "the flow through the taps backs up and slows down"). Or you have to build a bigger bath, which takes effort.

In a small but growing company (either by being in a brand new sector, or a new competitor with an advantage over the established ones), there's room for growth just by keeping the plug in.

But at some point, an investor wants to be able to get out more cash than they put in. With a no-dividend-but-growing company that's fine, as long as the market in it is still liquid (as it should be, in a bath ...). But what if there's a risk of the company's products becoming outdated? Reinvesting in a company builds up the company infrastructure, which is great if its products still sell, but having shares in the world's largest buggy-whip manufacturer is little consolation if you never received a dividend before those damn horseless carriages came along. Dividends are at least cash, rather than a specialised system that may not be efficiently convertible to making something else.

All true, except of course, you can dip the bucket into the bath at any point YOU choose, rather than wait expectantly beneath the plughole for the Directors to decide when to remove the plug, and how much water to let out. In the extreme case you use that disappointment at holding a non-dividend paying buggy whip manufacturer is mainly self-inflicted.

[Itsallaguess]

So if we purely concentrate on the bath analogy for now then, am I right in thinking that it's really only trying to describe two 'twin' businesses, where one of those business-twins reinvests it's own capital 'back into itself', and the other business-twin chooses to pay out more capital in the form of dividends, and where the subsequent re-investment of those dividends back into the same source business must then generate frictional costs that, ultimately, can only ever be at the detriment of the investor owning shares in that second dividend-paying business, compared to an investor owning shares in the first business?

Yes.

(Although I dislike absolute words like "only" and "must").

Thanks, and I agree, my previous use of the words 'only' and 'must' are probably unhelpful when hopefully asking someone to agree on a number of other related points, so thanks for pointing that out, but I'm also happy that we broadly agree on the above statement.

Moving on then, and trying to concentrate on the 'what happened' for now, without necessarily getting into any potential at this specific point for 'why it might have happened', are you happy that the Credit Suisse long-term study linked in my opening post shows that generally, and over the long-term, they show in their long-term US and UK market data that companies that pay out a high level of dividends (high yield) outperform on a total-return basis, those companies that pay out a low level of dividends (low yield), where they've regularly performed their yearly rebalancing processes as part of that long-term study?

Cheers,

Itsallaguess

[dealtn]

So if we purely concentrate on the bath analogy for now then, am I right in thinking that it's really only trying to describe two 'twin' businesses, where one of those business-twins reinvests it's own capital 'back into itself', and the other business-twin chooses to pay out more capital in the form of dividends, and where the subsequent re-investment of those dividends back into the same source business must then generate frictional costs that, ultimately, can only ever be at the detriment of the investor owning shares in that second dividend-paying business, compared to an investor owning shares in the first business?

Yes.

(Although I dislike absolute words like "only" and "must").

Thanks, and I agree, my previous use of the words 'only' and 'must' are probably unhelpful when hopefully asking someone to agree on a number of other related points, so thanks for pointing that out, but I'm also happy that we broadly agree on the above statement.

Moving on then, and trying to concentrate on the 'what happened' for now, without necessarily getting into any potential at this specific point for 'why it might have happened', are you happy that the Credit Suisse long-term study linked in my opening post shows that generally, and over the long-term, they show in their long-term US and UK market data that companies that pay out a high level of dividends (high yield) outperform on a total-return basis, those companies that pay out a low level of dividends (low yield), where they've regularly performed their yearly rebalancing processes as part of that long-term study?

Cheers,

Itsallaguess

I would say that study shows an investment strategy based on value investing which as a consequence generally involves investing in shares that initially have a high dividend yield, with subsequent "rebalancing" as part of that strategy, generally outperforms over the very long term on a total return basis. The study doesn't show that investing for dividends as an investment strategy outperforms.

I'm not sure why I am continuously being asked what the study says. The authors do that perfectly satisfactorily themselves - although again this is only a single paper of the large volume of work they have produced. Might your time not be better spent reading some of that if this is an area that interests you?

[Itsallaguess]

Moving on then, and trying to concentrate on the 'what happened' for now, without necessarily getting into any potential at this specific point for 'why it might have happened', are you happy that the Credit Suisse long-term study linked in my opening post shows that generally, and over the long-term, they show in their long-term US and UK market data that companies that pay out a high level of dividends (high yield) outperform on a total-return basis, those companies that pay out a low level of dividends (low yield), where they've regularly performed their yearly rebalancing processes as part of that long-term study?

I would say that study shows an investment strategy based on value investing which as a consequence generally involves investing in shares that initially have a high dividend yield, with subsequent "rebalancing" as part of that strategy, generally outperforms over the very long term on a total return basis.

The study doesn't show that investing for dividends as an investment strategy outperforms.

'Value investing' seems to not accurately describe what they initially set out to do though, even if it might be accurate to label it like that as part of the 'why it happened' discussion that might be considered separate to the point I was trying to make.

On the face of it though, they absolutely set out to use 'high yield' investments as the main primary starting point for their study, and also a primary driver for their regular rebalancing process, so if we take 'high yield' to be higher dividend-paying companies, can you please explain in more detail, taking this into account, what you mean when you say 'The study doesn't show that investing for dividends as an investment strategy outperforms'?

Please note that as I mentioned to mc2fool earlier, I am trying to concentrate this thread on the 'total-return pot-building' that was seen in the study, and I would prefer to completely ignore, for now, anything that might occur further down the line in terms of an investor eventually looking to actually remove any dividends from that pot.

I'm not sure why I am continuously being asked what the study says. The authors do that perfectly satisfactorily themselves - although again this is only a single paper of the large volume of work they have produced. Might your time not be better spent reading some of that if this is an area that interests you?

Apologies for what might feel like a continuation in questions. I would completely understand if you'd prefer not to put yourself in that position, but my opening post on this thread was clear that part of where I was hoping to gain a better understanding of things was connected and related to Rob's original dividend-related statement, your bath-analogy that attempted to refute it, and also the Credit Suisse study that seems to give some credence to the idea that good, long-term total-return results can be achieved by using high-yield as a driver for initial investment preference.

If we go back to what Rob said -

Dividends, growth in dividends and reinvested dividends are the main source of equity returns over the long run.

Given that Rob went on to cite Dimson and Siegel as providing evidence that he thinks shows the above statement is correct, and having now got a better understanding of at least this long-term Credit Suisse study that they were involved with, I am coming round to thinking that the underlying theme that Rob was trying to describe with the above statement is accurately reflected in the referenced Credit Suisse data, but that by wording his statement in the way that he has, it's left itself open to misinterpretation, where it perhaps can then be criticised using the type of 'twin company' bath analogy that you've come up with.

Do you think that's a fair statement that perhaps reflects this particular situation and what's led us to this point, where Rob might think that he's perhaps accurately describing what the Credit Suisse study finds, but wording his statement in a slightly inaccurate way, and also, as the above statement currently stands, it subsequently opens itself up to then being discredited with your bath-analogy?

I do appreciate the time and effort you've already put in to helping my understanding on this thread.

Cheers,

Itsallaguess

[dealtn]

I do appreciate the time and effort you've already put in to helping my understanding on this thread.

Firstly I didn't come up with a twin bath analogy.

I had a single bath representing (any) single company PURELY to explain it is what comes out of the taps, and doesn't escape, that delivers.

Any attempt to deliver multiple baths, and comparisons between them, be that better taps, more taps, more plugs, shinier taps etc. has been done subsequnetly, and by others. I have literally stated extended the analogy to multiple baths, and multiple market valuations was beyond its purpose

Secondly I agree "Value Investing" wasn't what they set out to do. But it is what they discovered as a reasonable explanation of what worked. And again if look at the library of their work they didn't set out to look at High Yield investing either.

Thirdly the analysis doesn't show that investing for dividends works. At best it shows that (initial) selecting using high dividend yield (actually not even that since "high" means above average) as a filter produces a portfolio with the potential to outperform. That outperformance isn't shown to be driven by the dividends received. As you correctly say it is the Total Return that matters which includes the change in the share price. Dimson et al go on to show that the relative discount in share price associated with such a selection, and its correction to no (or perhaps negative) discount in that share price delivers an outperformance (and it might be expected to be shown in reverse in an alternatively selected low, or no, dividend filtered portfolio, displaying a share price premium). The ideal Total Return of a value driven strategy, would be to buy at a discount, and sell at a premium, in such a short time period that no dividend was paid at all.

So

If we go back to what Rob said -

Dividends, growth in dividends and reinvested dividends are the main source of equity returns over the long run

1) Dividends. They come from earnings (at least in the long term) and those earnings could be retained. Even in the real world those dividends only work as a source of return if they are reinvested, in other words start as capital, transiently become income, and then turned back into capital again (which drive the earnings). Calling dividends the source of equity returns is wrong - unless you have a very peculiar definition of what "source" means.

2) Growth in Dividends. Again this growth isn't a source as it only comes on the back of (presumably) growth in earnings.

3) Reinvested Dividends. This is simply recycling as described - and has frictional costs, so less efficient than alternatives (though quite normal in the world of investment). However it clearly plays a very important role in delivering returns. Exactly the same as a savings account where interest accrues and is retained to compound delivering a much better final balance than alternatives where interest is paid out (and spent).

I find it frustrating that 2 different things are being conflated.

1) What drives equity returns.

2) Is there a "method" of delivering superior returns.

1) I think is self evident. It is the earnings of the underlying company, and more specifically that portion that is "owned" by the equity owners and not paid to other liability holders of the company.

2) Dimson et al have postulated, and I broadly agree, that an initial selection using above average dividend paying companies, can deliver that outperformance. BUT that outperformance is down to changes in valuation, not the receipt of dividends, despite their selection being determined by that dividend's existence. There is certainly correlation (at selection) but not causation. FURTHER the outperfomance, over a suitably long period of time, is driven by rebalancing (as the value is outed). Buy and hold of that initial selection, or harvesting the dividends from them, isn't the driver of the premium total return (but clearly an important component of the portfolios total return).

[simoan]

2) Dimson et al have postulated, and I broadly agree, that an initial selection using above average dividend paying companies, can deliver that outperformance. BUT that outperformance is down to changes in valuation, not the receipt of dividends, despite their selection being determined by that dividend's existence. There is certainly correlation (at selection) but not causation. FURTHER the outperfomance, over a suitably long period of time, is driven by rebalancing (as the value is outed). Buy and hold of that initial selection, or harvesting the dividends from them, isn't the driver of the premium total return (but clearly an important component of the portfolios total return).

I seriously couldn't agree more with this. IMHO it depends what is causing the high yield in the first place i.e. is it because the company is overstretching itself and paying out the entirety of its earnings as a dividend each year, or because the share price has fallen a long way? The dividend itself has nothing to do with subsequent performance and the return comes from an increase in valuation caused by one of the strongest forces in investment: "Reversion to the mean". In fact, dividend yield should not be studied standalone like this IMO, because many shares that have a high dividend will very likely also have one or more important value criteria i.e. Low P/E, Low P/CF and/or low P/B.

The first investment book I ever read (before I found TMF) was well-known value investor David Dreman's "Contrarian Investment Strategies" most of which is still highly relevant, despite its age, including Chapter 8: "Boosting Portfolio Profits" in which he found high dividend yield (Low P/D) was the least effective value criteria in driving returns. It's worth reading if you haven't. Of the four contrarian strategies studied, returns were best for Low P/E, then Low P/CF, then Low P/BV and last of all Low P/D in that order, over all time periods (5, 10, 15, 20 and 25 year).

All the best, Si

[mc2fool]

1) What drives equity returns.
:
1) I think is self evident. It is the earnings of the underlying company, and more specifically that portion that is "owned" by the equity owners and not paid to other liability holders of the company.

And valuations, which have a major effect on equity returns. Here's a now somewhat dated chart of the DJIA 1900-2004 vs P/E that illustrates just how much of an effect valuations have (on just a quick dig around I couldn't find a more up to date chart; I had this one bookmarked but the site has disappeared, however, thankfully, it's on the Wayback Machine...)

https://web.archive.org/web/20130606203849/http://www.minyanville.com/assets/catalog/products/00KAT-Dow100YrsMV.jpg

[Itsallaguess]

If we go back to what Rob said -

Dividends, growth in dividends and reinvested dividends are the main source of equity returns over the long run

1) Dividends. They come from earnings (at least in the long term) and those earnings could be retained. Even in the real world those dividends only work as a source of return if they are reinvested, in other words start as capital, transiently become income, and then turned back into capital again (which drive the earnings). Calling dividends the source of equity returns is wrong - unless you have a very peculiar definition of what "source" means.

2) Growth in Dividends. Again this growth isn't a source as it only comes on the back of (presumably) growth in earnings.

3) Reinvested Dividends. This is simply recycling as described - and has frictional costs, so less efficient than alternatives (though quite normal in the world of investment). However it clearly plays a very important role in delivering returns. Exactly the same as a savings account where interest accrues and is retained to compound delivering a much better final balance than alternatives where interest is paid out (and spent).

I find it frustrating that 2 different things are being conflated.

1) What drives equity returns.

2) Is there a "method" of delivering superior returns.

1) I think is self evident. It is the earnings of the underlying company, and more specifically that portion that is "owned" by the equity owners and not paid to other liability holders of the company.

2) Dimson et al have postulated, and I broadly agree, that an initial selection using above average dividend paying companies, can deliver that outperformance. BUT that outperformance is down to changes in valuation, not the receipt of dividends, despite their selection being determined by that dividend's existence. There is certainly correlation (at selection) but not causation. FURTHER the outperfomance, over a suitably long period of time, is driven by rebalancing (as the value is outed). Buy and hold of that initial selection, or harvesting the dividends from them, isn't the driver of the premium total return (but clearly an important component of the portfolios total return).

I agree, but with the caveat that mcfool2 has already raised, which is to highlight the large influence that different valuations also have on long-term returns, and I think we've seen in this thread already that there's often a real focus on 'dividends' and their potential to influence returns (or not..), due to a quite natural tendency to concentrate on the 'high-yield' nature of the debate, whilst perhaps forgetting to also remember that two different baths are perhaps being 'rated' by the market differently from the outset, and so any frictional-costs associated with one of them might well be countered by valuation benefits of that high-yield bath and also a potential overvaluing of the low-yield alternative...

So I'm back to thinking that Rob was actually failing to properly articulate in his statement just what the Credit Suisse results were showing in the dividend-centric wording that he's used to do so.

With that said, I also think though, that we should again highlight what you've already said regarding the bath analogy, which is that it's only trying to show the frictional-cost nature of companies paying out dividends and any subsequent re-investment of them, and I think you've now been happy to make clear that it was never intended to cover any 'valuation difference' between any two given baths, where completely different market valuations might be placed on each of them, and which may (as shown in the Credit Suisse report) subsequently go on to play a large role in long-term total returns from investments in each of them.

If there are any glaring errors in the above summary then they are completely unintentional and I'd be keen to be corrected, as I'm certainly not trying to drive to any conclusions here, and am very much looking for a general consensus in pulling everything together in broad agreement on what's been a very interesting discussion.

On a personal level, I think it's unfortunate that these sorts of studies end up actually highlighting a level of 'superior returns', because doing so seems to then set a very high bar and real focus on subsequent debate around them. Personally, I would never seek to place 'superior returns' high on my investment agenda, as I have separate goals where I'm happy to give us a level of return if those other influences can play a part in the deliver of what I have always been happy to take as 'good-enough' returns.

Cheers,

Itsallaguess

[mc2fool]

It is, perhaps, worth gently reminding that, as I'm sure everyone knows, yield is not the same as dividends. Some companies can have a relatively low dividend payout (to their earnings, i.e. high cover) but a high yield, and vice-versa.

And while the DMS study (and HYP) select by yield (but then have a very different strategy going forward), Rob's fund, as he keeps on getting mentioned, OTOH doesn't invest by yield but by total £millions of dividends paid by the companies, so his fund has e.g. Diageo as its 11th largest holding even though it only yields 2%.

End of gentle reminder....

As dealtn keeps on mentioning, Dimson et al have done lots of research on a variety of matters and some of that is in the Credit Suisse Global Investment Returns Yearbooks, to which DMS have been supplying research since 2009, and they can all be found at:

https://www.credit-suisse.com/about-us/en/reports-research/studies-publications.html

They're all worth a read, IMO, although the more recent ones are "summary" editions containing extracts from the full copies of the yearbooks, which are free to CS clients and several hundred quid to anyone else! The summaries do, however, still contain several mugs of coffee's worth of reading. And for the 2022 yearbook there's also a 50 video minute presentation by Dimson & Marsh (see link above and select the 2022 yearbook).

In one of the initial announcements (which I can't find now) for the 2011 one it was mentioned that DMS had found premiums for momentum and size (smaller), as well as yield. Here's some other research papers that, although a little dated, may be of interest:

108 Years of Momentum Profits Dimson, Marsh & Staunton. 2008.
Capturing the Value Premium in the United Kingdom Dimson, Nagel & Quigley. 2003.
Surprise! Higher Dividends = Higher Earnings Growth Arnott & Asness. 2003

[Itsallaguess]

Surprise! Higher Dividends = Higher Earnings Growth Arnott & Asness. 2003

From the first page -

We investigate whether dividend policy, as observed in the payout ratio of the U.S. equity market portfolio, forecasts future aggregate earnings growth.

The historical evidence strongly suggests that expected future earnings growth is fastest when current payout ratios are high and slowest when payout ratios are low.

This relationship is not subsumed by other factors, such as simple mean reversion in earnings.

Our evidence thus contradicts the views of many who believe that substantial reinvestment of retained earnings will fuel faster future earnings growth.

[gulp]

Cheers,

Itsallaguess

[Lootman]

Rob's fund, as he keeps on getting mentioned, OTOH doesn't invest by yield but by total £millions of dividends paid by the companies, so his fund has e.g. Diageo as its 11th largest holding even though it only yields 2%.

I believe that his fund actually weights by expected/anticipated dividends rather than actual dividends paid. This is done by using analysts' predictions of dividends in the next 12 months rather than dividends paid in the last 12 months:

The Fund aims to maximise overall returns by seeking to replicate the performance of the Elston Smart-Beta UK Dividend Index.

The fund tracks an index whose security selection process captures changes in forecast dividends; weights holdings by forecast dividend contribution; and adjusts for dual-listings

https://www.valu-trac.com/administratio ... nts/munro/

Your point about dividends being quite different from yield is well taken. A good example is Apple, which yields only 0.6% but, because of its huge market capitalisation, that equates to a payout of nearly $15 billion a year!

So Ohno/Munro would give it a high weighting but a HY approach would ignore it.

[Charlottesquare]

To anyone who has actually read the Credit Suisse study, if the study envisages selling each year and then buying the next year's prime candidates, how do they account for the frictional costs of this buying and selling each year when arriving at their return figures, have they included these costs?