A warning on dividend investing

[Gosso]

stone,

Aren't you ignoring the fact that some growth companies will evolve into dividend companies.  Also growth companies can be gobbled up by other growth or dividend companies.  I have no idea what the ratio of failure/evolution/acquisition is, but they don't all eventually burn out and die.

I think dividend investing can work, but it does so because it focuses on mature companies with a steady revenue.  There is nothing wrong with that (it is very Buffett-like), but dividends have very little to do with the success, IMO.

[stone]

stone,
Aren't you ignoring the fact that some growth companies will evolve into dividend companies.  Also growth companies can be gobbled up by other growth or dividend companies.  I have no idea what the ratio of failure/evolution/acquisition is, but they don't all eventually burn out and die.
I think dividend investing can work, but it does so because it focuses on mature companies with a steady revenue.  There is nothing wrong with that (it is very Buffett-like), but dividends have very little to do with the success, IMO.

Some non-dividend companies do evolve into dividend companies (and vice versa) but I got the impression that Sophie wasn't advocating holding a non-dividend stock with a view to later getting a stream of dividends from it but rather simply holding non-dividend companies in the belief that the share price would exhibit durable increases over time. Historically stocks have paid out above and beyond inflation and the rate of economic growth. It would be impossible for them to do that whilst retaining all of that value as a durable increase in stock price- that would create an anomaly because they would immediately become overvalued and that in its self would prevent them from being able to perform like that. The only ways to fix that are either to convert that gathering in of money into dividends OR to convert it into price gyrations. I think stock price gyrations are a big blind spot for some people.

There is something in the link below explaining how an economic absurdity would occur if stock returns were in the form of durable stock price increases:
http://www.gmo.com/America/  Reports of the Death of Equities Have Been Greatly Exaggerated:
Explaining Equity Returns

The gap between the 1.7% real earnings growth (about half the rate of GDP growth) and 5.9% real return (almost
double the rate of GDP growth) is made up by dividends, which have averaged about 3.9% since 1929, and a bit
of valuation shift (the P/E of the market is a couple of points higher today than it was in December of 1929). So if
aggregate market capitalization has grown along with GDP and the compound return to equities has been much faster,
what gives? Do those original shareholders control 8 times as much of economic output as they did 81 years ago?
Of course they don’t. Investors don’t invest to simply accumulate wealth that is never to be spent. Workers invest
to fund their retirements. Pension funds and insurance companies are obligated to service their required payouts.
Endowments and foundations pay out 5% or so of their total value every year to fund the causes and organizations

[AgAuMoney]

Stocks decline by the amount of their dividend after the ex dividend date.

That's true.  They do adjust the previous day close by the amount of the dividend.

But then the stocks also decline more or sometimes they go up as soon as the market opens.

In other words, for the typical dividend, there is zero net impact on the stock price.

Huh????  The logic of this completely escapes me.  What you mean is that the stock price decrement is hidden by the normal fluctuation, but that doesn't mean it's not there.

The logic is that although there is an accounting change to the price, that change is of zero effect in the real world.

It is the difference between theory and practice.  In theory there is no difference.

In this case the theory is that there is a difference, a negative difference.

In reality there is no difference.  The normal, real, flucuations of the market completely swamp any effect of the price change.  You cannot tell by looking only at prices that a dividend adjustment was or was not made.  There is no difference.

[stone]

I'd further that the "theory" is total bunk and that is why it fails to reflect reality. Even as a theory it is internally inconsistent unless you were to imagine that all companies were endlessly expandable conglomerates such as Berkshire Hathaway. Of course they aren't. They almost always have a limited niche of profitability and they can't increase in value without becoming overvalued -so the actual choice is between dividends and price gyrations not between dividends and endless expansion.

[AgAuMoney]

I think dividend investing can work, but it does so because it focuses on mature companies with a steady revenue.  There is nothing wrong with that (it is very Buffett-like), but dividends have very little to do with the success, IMO.

It is hard to separate cause from effect.

There are multiple studies from multiple sources covering multiple markets (primarily US and western Europe) that show dividends provide most of the total return of the market.  In other words, price changes (cap gains) provide less than half of what we could consider the market return.

Now is that due to the dividends actually being paid out?  Or is that due to the nature of the companies that do pay dividends?

Hard to say for certain.

My opinion is that being a good company allows them to pay dividends.  And the dividends are the reason for the return.  Neither one works without the other.

Further, after more than 10 years of debating it with dozens of people I've learned that capital gains do not compound.

Repeat:  Cap gains do not compound.

It's a mathematical truism because gains can come and go.

If you have a gain YOU can compound it if you sell and invest in something else that gains.  But if you are sitting on the gain or if you put the gain into something else and you lose 40% like in 2008, future gains are NOT based on the high value from before, but from the new low value.  That is not compounding.

There is no stake in the ground that they cannot take back.  That's not compounding.

Dividend growth does compound.

When a company increases their dividend 10% (and many companies have decades of paying more than the year before) it is an increase over the year before.  The next year when they increase their payout it is compounded on the previous increase.

That is compounding.

If a company threatens to or stops growing their dividend then that compounding is at risk, and you should consider taking action.

Unlike gains, with dividends they cannot take back what they have already paid.

And it gets better, because owning shares of dividend growth companies entitles you to future payments in a fixed amount relative to each share.  The price of the shares doesn't matter.  So when you get more shares, you get more dividends.  If you use the dividends to buy more shares, the next time you get more dividends.  That is another layer of compounding.  Your dividend payments go up every quarter.  And then when the company increases their dividend, your payment goes up even more.

The effect of those two layers of compounding feeding each other over the past 10 years has been incredible for my portfolio.

Compounding is truly a wonder, and you don't get it if all you are doing is playing for cap gains unless you drive it yourself and manage to avoid the market taking it back.

[Gumby]

For what it's worth, Wikipedia has a list of criticisms on dividend investing...

Some believe that company profits are best re-invested back into the company: research and development, capital investment, expansion, etc. Proponents of this view (and thus critics of dividends per se) suggest that an eagerness to return profits to shareholders may indicate the management having run out of good ideas for the future of the company. Some studies, however, have demonstrated that companies that pay dividends have higher earnings growth, suggesting that dividend payments may be evidence of confidence in earnings growth and sufficient profitability to fund future expansion.

Taxation of dividends is often used as justification for retaining earnings, or for performing a stock buyback, in which the company buys back stock, thereby increasing the value of the stock left outstanding.

When dividends are paid, individual shareholders in many countries suffer from double taxation of those dividends:

 1) the company pays income tax to the government when it earns any income, and then
 2) when the dividend is paid, the individual shareholder pays income tax on the dividend payment.

In many countries, the tax rate on dividend income is lower than for other forms of income to compensate for tax paid at the corporate level.

Capital gains should not be confused with dividends. Capital gains assumes an increase in a stock's value. Dividend is merely parsing out a share of the profits, and is taxed at the capital gains tax rate. If there is an increase of value of stock, and a shareholder chooses to sell the stock, the shareholder will pay a tax on capital gains (often taxed at a lower rate than ordinary income). If a holder of the stock chooses to not participate in the buyback, the price of the holder's shares could rise (as well as it could fall), but the tax on these gains is delayed until the actual sale of the shares.

Certain types of specialized investment companies (such as a REIT in the U.S.) allow the shareholder to partially or fully avoid double taxation of dividends.

Shareholders in companies that pay little or no cash dividends can reap the benefit of the company's profits when they sell their shareholding, or when a company is wound down and all assets liquidated and distributed amongst shareholders. This, in effect, delegates the dividend policy from the board to the individual shareholder. Payment of a dividend can increase the borrowing requirement, or leverage, of a company.

Source: http://en.wikipedia.org/wiki/Dividend

Interestingly, it doesn't mention anything about compounding dividends as being a special benefit to investors.

[Gosso]

Ag and stone,

I'm not suggesting that one should avoid companies that pay a dividend, but I am suggesting that one shouldn't avoid companies that don't pay a dividend.

Where does the money to pay the dividend come from?  It comes from the cash on the balance sheet, which decreases the book value of that company.  Of course the stock price is going to gyrate, but that doesn't change the fact that the company's value has decreased by the amount of the dividend distributed.

My simple mind cannot understand the compounding of dividends vs capital gains...too many mental gymnastics for me

[stone]

Ag and stone,

I'm not suggesting that one should avoid companies that pay a dividend, but I am suggesting that one shouldn't avoid companies that don't pay a dividend.

Where does the money to pay the dividend come from?  It comes from the cash on the balance sheet, which decreases the book value of that company.  Of course the stock price is going to gyrate, but that doesn't change the fact that the company's value has decreased by the amount of the dividend distributed.

My simple mind cannot understand the compounding of dividends vs capital gains...too many mental gymnastics for me

Gosso, just imagine if the stock market dividends from the last 100 years had been kept by companies and successfully compounded over that time. So rather than paying a 6% dividend, each company had instead kept that money and somehow managed to successfully deploy it to the same extent as they deployed the book value that they actually had over the past 100 years. That hypothetical dividend-free history would have created an economic absurdity because the compounded growth would have created an absurdly high total stock market capitalization. The economy would have become entirely unable to provide the cash flow to maintain the P/E of that vast stock price. Sorry if I'm missing something and just in a muddle about this but I really don't think I am.

[Gosso]

Ag and stone,

I'm not suggesting that one should avoid companies that pay a dividend, but I am suggesting that one shouldn't avoid companies that don't pay a dividend.

Where does the money to pay the dividend come from?  It comes from the cash on the balance sheet, which decreases the book value of that company.  Of course the stock price is going to gyrate, but that doesn't change the fact that the company's value has decreased by the amount of the dividend distributed.

My simple mind cannot understand the compounding of dividends vs capital gains...too many mental gymnastics for me

Gosso, just imagine if the stock market dividends from the last 100 years had been kept by companies and successfully compounded over that time. So rather than paying a 6% dividend, each company had instead kept that money and somehow managed to successfully deploy it to the same extent as they deployed the book value that they actually had over the past 100 years. That hypothetical dividend-free history would have created an economic absurdity because the compounded growth would have created an absurdly high total stock market capitalization. The economy would have become entirely unable to provide the cash flow to maintain the P/E of that vast stock price. Sorry if I'm missing something and just in a muddle about this but I really don't think I am.

I agree that at some point companies need to do something with their excess cash.  It doesn't make any sense for it to build-up (think Apple).  So they can pay a dividend, buy back shares, invest in growth/R&D or look for good investment/acquisition targets.  In some/most cases the best option is to pay a dividend.  This decreases their excess cash, and their book value (which the market should adjust for by gyrating their stock at a lower level).

[stone]

Gosso, I wonder whether a 3xLTT ETF might be analogous to many non-dividend companies. With that ETF, all of the bond coupons go to pay the borrowing costs so the 3xETF holder is left with just the volatility. Similarly with many companies, much of the cash flow goes to the bond-holders (or bank loans) and the share holders are left with volatility. Being a buy-and-hold investor with a 3xLTT ETF would be just plain dumb and yet perhaps many people do just the same thing with some of the stocks they hold

[stone]

Clive, I agree that in principle borrowing could be used to actually increase the dividend yield for the shareholders as you describe but in practice if you compare very similar companies with different levels of borrowing, then what typically seems to actually occur is that borrowing causes dividends to be replaced by greater volatility just like with a 3xLTT ETF (as in that example of comparing Rio Tinto with BHP Billition that I mentioned before).
Like you say, volatility can be captured and so turned into a cash flow for the shareholder BUT ONLY if the shareholder goes about doing that.

[Gosso]

Gosso, I wonder whether a 3xLTT ETF might be analogous to many non-dividend companies. With that ETF, all of the bond coupons go to pay the borrowing costs so the 3xETF holder is left with just the volatility. Similarly with many companies, much of the cash flow goes to the bond-holders (or bank loans) and the share holders are left with volatility. Being a buy-and-hold investor with a 3xLTT ETF would be just plain dumb and yet perhaps many people do just the same thing with some of the stocks they hold

Correct me if I'm wrong but I think the fundamental difference between our arguments is that you're assuming most growth stocks will go to zero (ie RIM, Nortel), and that the shareholders will have nothing to show for their investment, and would have been better off if RIM/Nortel had paid a dividend.  I agree with this.  But, I'm assuming that most growth companies evolve into dividend companies, or are acquired by bigger fish.  I wonder what the ratio is...

Most growth ETF's perform fairly well (although typically not as well as value), so if one holds enough of these growth companies then they produce returns somewhat similar to the S&P 500.

[Gosso]

Clive, I agree that in principle borrowing could be used to actually increase the dividend yield for the shareholders as you describe but in practice if you compare very similar companies with different levels of borrowing, then what typically seems to actually occur is that borrowing causes dividends to be replaced by greater volatility just like with a 3xLTT ETF (as in that example of comparing Rio Tinto with BHP Billition that I mentioned before).
Like you say, volatility can be captured and so turned into a cash flow for the shareholder BUT ONLY if the shareholder goes about doing that.

Stocks that pay dividends are more like a stock/bond blend, whilst stocks that retain all earnings are more pure all-stock like. Pure stocks will tend to be more volatile because of relatively more being exposed to that single asset class. Stock-bond combinations tend to have lower volatility.

This makes sense.  I think I now understand what stone meant when discussing the volatility of non-dividend stocks.

[AgAuMoney]

Doesn't Warren Buffett say that he wont pay dividends because he likely can earn a higher return on reinvestment of dividends than what the dividend receiver might.

That's what he says.

And for much (but not all) 5+ year periods since 1970 that has been true for Berk when taken as a whole.

But dividends are not the whole.  If you look at JUST the cash sitting in Berk, the numbers and the math get rather messy trying to figure everything out but it's reasonably clear.  With Apple it became painfully obvious that the cash inside the company was NOT generating good returns.  How can it, the company is holding it in CASH which pays DIDDLY!  (Multiply the size of the cash by any reasonable return figure for cash and compare that with the return you would have received if they had distributed 25% of the cash to shareholders and you used your distribution to buy more shares at a price reduced by the dollar amount per share of the distribution.  Since they never have used anywhere near 75% of their cash, such a reduction would not have any effect on their other returns.)

If you read a few Warren bio's you will realize that the reason he sits on a huge cash hoard instead of paying out a dividend is because he is simply a greedy miser and it's all about the control and the security of having all that money.  Once he has his hands on money it hurts him to let it go so he doesn't until he is convinced the distribution is necessary or for the best.  He treated his family much the same way until his kids were grown and he was finally convinced to give them some money.

[stone]

Berkshire Hathaway is also quite different from most companies in that it is a conglomerate and so can "effortlessly" make use of retained earnings by simply buying more shares in other companies. Apple didn't have that effortless outlet for using retained earnings. Am I right in understanding that the traditional way for large tech companies to use retained earnings is to buy-up and snuff out small potential competitors? I guess the big tech companies are basically paying a ransom to avoid competition from the people creating new-upstarts. Another type of non-dividend companies are the debt-wrecks that have been wrung out by Romney style private equity. After a leveraged buy-out followed by a period of having all cash flow directed to debt servicing, the resulting husk of a company is then returned to the market.

I do think though that the BHP Billiton versus RioTinto comparison distills down the essence of a very prevelent phenomenon of a trade off between volatility and dividends. Both of those companies are very long established iron ore miners. It isn't a comparison between an established company and an upstart or anything like that. The difference between them is basically down to RioTinto having lots of borrowing. You could choose to make RioTinto shares act like BHP Billiton shares by holding them together with cash and rebalancing or you could make BHP Billiton shares act like  Rio Tinto shares by borrowing money to buy them and using the dividends to pay the borrowing cost. Basically it is leverage internalized into the stock just as you get with a leveraged ETF. The key thing is to see it for what it is. You'd be very mistaken if you thought that it made sense to follow a buy and hold strategy with Rio Tinto waiting for the day when it produced a stream of dividends (or durable exponential expansion of the share price as per Berkshire Hathaway).

[stone]

I really don't think that a company with lots of internal leverage (such as with Rio Tinto) is comparable to a zero bond. To me it is much more comparable to a 3xLTT ETF. Both a zero bond and a 3xLTT ETF are  securities that pay no income BUT the zero pays out a guarenteed amount at maturity. The zero has a locked in capital gain. If you spend $1 today on a zero and hold it for 30years then you get a payment of $2 or whatever it is. The 3xLTT ETF does no such thing- it just gyrates around indefinately and the income from the underlying treasuries goes to pay the internal borrowing costs of the leveraged ETF. Clive, you know that better than anyone and you are always coming up with ingenious ways to harvest the price gyrations of 3xLTT ETFs.
Rio Tinto is no upstart growth stock, it is a very very long established "forever stock". My understanding is that it intends to be around in a hundred years time, still owning 20% of the world's iron ore reserves (or whatever it is), still not paying a significant dividend and with a stock price that swings wildly to and fro but that certainly does not follow the impossible course of continuous exponential growth. Let's face it, that is also the intended future for the 3xLTT ETF.
I guess canny, activist, shareholders (and insiders) much prefer to have price gyrations than dividends because price gyrations are there for the taking but only by those who manage to harvest them. It is a sleight of hand whereby returns can bypass the "dumb money" and all get taken by those who harvest volatility.

[AgAuMoney]

Berkshire Hathaway is also quite different from most companies in that it is a conglomerate and so can "effortlessly" make use of retained earnings by simply buying more shares in other companies. Apple didn't have that effortless outlet...

Berk is only a conglomerate BECAUSE they started buying shares in and entire other companies.  They were a cloth mill when Warren Buffet bought them.

Apple could also buy stock in other companies or buy them outright.

It would earn them the scorn of investors everywhere, and probably people who weren't investors and I imagine their board of directors would quickly put a stop to it.  But there is no technical or legal reason why Apple limits themselves.

Buffet specifically folding his investing partnership and kept one holding (Berk) specifically so he could run it any way he pleased (with his majority ownership) as a public company with the freedom that permits, without all the restrictions, disclosure requirements and demands imposed on an investing partnership.

[WildAboutHarry]

Repeat:  Cap gains do not compound.

I'm having a hard time with this concept.  Can you elaborate or point me to a source to review the concept?

Assuming two identical companies, one pays dividends, one doesn't.  Assuming you reinvest the dividends, don't you end up in the same place?  That is, the value of the stock plus reinvested dividends (first company) should match the value of the stock (second company).

[stone]

Repeat:  Cap gains do not compound.

I'm having a hard time with this concept.  Can you elaborate or point me to a source to review the concept?

Assuming two identical companies, one pays dividends, one doesn't.  Assuming you reinvest the dividends, don't you end up in the same place?  That is, the value of the stock plus reinvested dividends (first company) should match the value of the stock (second company).

I think that that is only true for a situation such as Berkshire Hathaway where the retained earnings can be used to buy more of the same (ie more shares in other companies in the Berkshire Hathaway case). It is much more typical for a company to have a limited niche such that it has an optimal size and it is very wasteful and unproductive for it to try and expand beyond that niche. You can see that when small companies get aquired by big ones and become dysfunctional.
Above and beyond that business model issue is the issue with leverage. A company can borrow money to the extent that almost all cash flow goes towards servicing the company's debt. In many cases your non-dividend paying stock is using retained earnings to pay down debt and then when times are good using the increased cash flow and stock price to take on more debt. Basically the stock is acting like a leveraged ETF and just like a leveraged ETF it has amplified volatility. That is great for those who buy low and sell high but any potential gains will entirely pass by a buy and hold share owner.

[WildAboutHarry]

Stone,

I see the point about leverage and realizing gains and avoiding losses, but isn't that is just market timing on steroids?

I looked at a bunch of Vanguard's Dividend Strategy funds (Dividend Growth, Equity Income, et al.) compared to the Total Stock Market fund.  Not a lot of performance difference.  Is TSM just the weak form of a closet dividend strategy fund?

After reading this thread I confess to Googling about dividend strategies and felt a nostalgic longing for owing individual stocks.  KO! WAG! XOM! MCD!  They are all there! 

[MachineGhost]

See the pretty charts in this thread: http://gyroscopicinvesting.com/forum/ht ... ic.php?t=1

[stone]

Stone,
I see the point about leverage and realizing gains and avoiding losses, but isn't that is just market timing on steroids?
I looked at a bunch of Vanguard's Dividend Strategy funds (Dividend Growth, Equity Income, et al.) compared to the Total Stock Market fund.  Not a lot of performance difference.  Is TSM just the weak form of a closet dividend strategy fund?
After reading this thread I confess to Googling about dividend strategies and felt a nostalgic longing for owing individual stocks.  KO! WAG! XOM! MCD!  They are all there!  

I don't think volatility capture requires market timing. A rebalancing strategy with fixed proportions of each holding is a volatility capture strategy without market timing. A covered call option strategy is another. I think craigr's argument for holding just TSM in a HBPP is that he considers that the bulk of stock volatility is synchronised across all stocks and so gets captured by the HBPP (sorry if I've misunderstood). I still suspect that the capital structure of many stocks is set up specifically so that gains wriggle past buy-and-hold share holders. Definately be very wary of holding stocks that have internal leverage within a VP without any sort of volatility capture. I don't think it is sensible to suppose that they will "grow" in any long term sense.
When you compared the dividend stocks to the TSM, were you reinvesting the dividends?  I guess that that has an effect that compounds considerably over time even though it is small in the short term (as AuAgmoney described).

[WildAboutHarry]

When you compared the dividend stocks to the TSM, were you reinvesting the dividends?  I guess that that has an effect that compounds considerably over time even though it is small in the short term (as AuAgmoney described).

Yes.  I used Morningstar and ran the comparison chart with the "growth" setting which shows reinvested dividends.  Of course the results are period-dependent, but TSM tracks most of the Vanguard dividend funds reasonably closely.  The dividend funds held up relatively well in 2008 (compared to TSM).  And they do "outgrow" TSM over long terms.

The yield of TSM is about 2% and most of the Vanguard dividend funds are under 3%.  If TSM is not just a low-power dividend strategy fund, how does one "compound" the capital gains?  Or is that not possible?

See the pretty charts in this thread:

Thanks MG, I missed this one.  I'll get a large cup of coffee and read it.

[stone]

WildaboutHarry

If TSM is not just a low-power dividend strategy fund, how does one "compound" the capital gains?  Or is that not possible?

To the extent that the volatility of different stocks is correlated, the HBPP strategy captures that volatility and allows it to be reinvested in a compounding way. So as stock prices bob up and down, they get rebalanced into gold and LTT and cash on the ups and back into stocks on the downs. I guess though that a HBPP with a TSM fund for the stocks must miss the volatility between individual stocks.

[WildAboutHarry]

So compounding capital gains from TSM, at least in the context of the HBPP, requires sufficient volatility to trigger re-balancing out of TSM and into something else and vice versa.  Otherwise, TSM is a low-power dividend strategy fund.

My recollection is that there have been few re-balancing events in the HBPP over the past forty years or so (Clive had a table, I think).  So the opportunities for compounding capital gains seem limited.