Is diversification a sham? (financial advice)

[Jeff Rosenbury]

What does that even mean? I agree it sounds very wrong.

It means that there will someday be a disaster that forces the value of the stocks on the market to zero (or close enough to not matter.) See my Data General example.

Wars, government seizures, disasters, legal status changes, tax code changes, economic innovation: These can all eliminate the value of stocks for investors. Historical markets went to zero. (The Tulip Market is an example.) They went to zero fairly quickly compared to the New York Stock Exchange which has an awesome track record.

But we are in a rapidly changing economy. The values of goods and services are shifting rapidly. Expecting traditional market stability seems short sighted to me.

Of course you can watch your investments carefully and avoid most of the downside associated with these sorts of events, but that is not the investment strategy favored by most saving for retirement. Long term diversified buy and hold (with periodic balancing) seems to work best for most pension planners. They are busy with their careers and don't really have the time and energy to watch every potential event that might affect their stocks.

Some might argue that those who don't watch their stocks deserve what they get, but that is unrealistic on at least two levels. First, I want my plumber working on my plumbing, not watching the market. Second, if his retirement dries up and blows away along with most of the rest of his age cohort, we all know we, the public, will bail him out. No, it is in the public interest to keep long term investments safe. One good way of doing that is to ensure dividends by investing in dividend paying stocks.

 

[BWV]

if you had bought and maintained a cap-weighted portfolio of all the world's publicly traded equity markets beginning in 1900 you would have earned an annualized return, in U.S. Dollars and net of inflation, of 5.4%, compared to 1.9% for a similar portfolio of government bonds. (http://www.investering.dk/documents/10655/157815/Credit+Suisse/0c0ceeb5-c5af-464d-be00-fb1378cb0412).

This includes a few markets that went to zero, such as Russia in 1917 and long interruptions in Japan and Germany during WW2. Only a bare handful of companies that existed in 1900 survived to the present and it is hard to imagine more wrenching disruption in the future than occurred during the 20th century (other than some apocalypse in which case investment decisions would not matter). So at least the history of the 20th century does not support your contention there

 

[andrewkirk]

other than some apocalypse in which case investment decisions would not matter

Investment decisions will not matter after the apocalypse. But it's the potential for an apocalypse that makes investment decisions before the apocalypse matter. If one receive dividends then one can use them to promote the happiness of oneself and others. If one receives no divs (or substitutes like buybacks), one cannot, so the pre-apocalypse years will have had less enjoyment.

The Miller-Modigliani theory is not in conflict with this, because it is predicated on an assumption of stable growth, so its application is necessarily limited to the short-to-medium term.

I find it very easy to imagine more wrenching disruption than occurred in the 20th century. Almost every previous century had more wrenching disruption than the 20th. We just focus on the 20th more because it's more recent, and because we have photos, video and audio of the events. Stephen Pinker's book 'The better angels of our nature' has some interesting analysis of this sort of thing.

 

[BWV]

Investment decisions will not matter after the apocalypse. But it's the potential for an apocalypse that makes investment decisions before the apocalypse matter. If one receive dividends then one can use them to promote the happiness of oneself and others. If one receives no divs (or substitutes like buybacks), one cannot, so the pre-apocalypse years will have had less enjoyment.

The Miller-Modigliani theory is not in conflict with this, because it is predicated on an assumption of stable growth, so its application is necessarily limited to the short-to-medium term.

I find it very easy to imagine more wrenching disruption than occurred in the 20th century. Almost every previous century had more wrenching disruption than the 20th. We just focus on the 20th more because it's more recent, and because we have photos, video and audio of the events. Stephen Pinker's book 'The better angels of our nature' has some interesting analysis of this sort of thing.

But according to MM one can always sell shares to raise cash. You are reciting what MM called the 'bird in the hand fallacy'. Corporate finance 101 is that companies should only pay dividends if they do not have investment opportunities where the return exceeds their cost of capital.

 

[andrewkirk]

But according to MM one can always sell shares to raise cash.

Let's focus on the MM theory then.

The theory says that, after making a bunch of very strong assumptions, the value to an investor of an investment in a company should not be affected by leverage, funding or dividend policy.

How is that value calculated?

 

[BWV]

Present value of cash flows, utilizing an appropriate discount rate. as long as a company has investment opportunities where the return exceeds its cost of capital this value is maximized by reinvesting earnings rather than paying them out as dividends

 

[andrewkirk]

What economic theory says that the value to an investor of an asset is the present value of its undistributed cash flows?

 

[Jeff Rosenbury]

if you had bought and maintained a cap-weighted portfolio of all the world's publicly traded equity markets beginning in 1900 you would have earned an annualized return, in U.S. Dollars and net of inflation, of 5.4%, compared to 1.9% for a similar portfolio of government bonds. (http://www.investering.dk/documents/10655/157815/Credit+Suisse/0c0ceeb5-c5af-464d-be00-fb1378cb0412).

This includes a few markets that went to zero, such as Russia in 1917 and long interruptions in Japan and Germany during WW2. Only a bare handful of companies that existed in 1900 survived to the present and it is hard to imagine more wrenching disruption in the future than occurred [sic] during the 20th century (other than some apocalypse in which case investment decisions would not matter). So at least the history of the 20th century does not support your contention there

That was a nice report.

I was discussing individual markets over the long term. From your report (p22), "We deem Venezuela, which was removed on 20 June 2003 with a total return index value on the 19th of 94.78 and on the 20th of zero, to have lost 99.99% of its value." As we see here, an entire market was destroyed overnight. The owners (stockholders) were dispossessed. Anyone planning on retiring had their savings eliminated. M&M theory broke down because no one could sell their stocks.

I am not arguing markets are bad investments. I am not arguing against the M&M theory for many investors. I am arguing that as retirement draws near, dividend paying stocks tend to be more stable, thus a good choice.

Modigliani-Miller theorem posits no taxes, bankruptcy costs, or asymmetric information. These are at best ideals to be achieved, not the reality on the ground.

 

[BWV]

That was a nice report.

I was discussing individual markets over the long term. From your report (p22), "We deem Venezuela, which was removed on 20 June 2003 with a total return index value on the 19th of 94.78 and on the 20th of zero, to have lost 99.99% of its value." As we see here, an entire market was destroyed overnight. The owners (stockholders) were dispossessed. Anyone planning on retiring had their savings eliminated. M&M theory broke down because no one could sell their stocks.

I am not arguing markets are bad investments. I am not arguing against the M&M theory for many investors. I am arguing that as retirement draws near, dividend paying stocks tend to be more stable, thus a good choice.

Modigliani-Miller theorem posits no taxes, bankruptcy costs, or asymmetric information. These are at best ideals to be achieved, not the reality on the ground.

In Venezuela or 1917 Russia, investors were screwed whether or not their stocks paid dividends
For an investor saving for retirement outside of a tax-deferred vehicle like a 401K dividend paying stocks are disadvantaged as the investor faces high marginal tax brackets on the additional income whereas companies that don't pay dividends can compound that growth internally, deferring taxes until a later date
practically, the issue of dividends really hinges on the maturity of the business and its growth prospects. The reason Cisco, Microsoft and other companies now pay dividends is that they are large mature businesses and cannot reinvest all their cash flow into high return projects.

 

[BWV]

What economic theory says that the value to an investor of an asset is the present value of its undistributed cash flows?

Just the plain old dividend discount model, but not sure where you got 'present value of undistributed cash flows' out of my earlier post.
So for a stock with price P with dividend d, growth rate g and discount rate r

P= ∑d(1+g)ⁿ/(1+r)ⁿ for i=1:n
using the definition of the limit of a geometric series, this converges at n→∞ to:
P=d/(r-g)
now this becomes nonsense if g≥r, but that is OK as there are economic arguments on how discount rates must tie to economic growth and growth rates converge over time to overall economic growth (obviously a company cannot grow its earnings faster than the economy forever). Practically, a 2 stage model is used to capture a shorter term high growth phase where g>r, you could also use a logistic model

this can be rewritten as d/P(dividend yield)=r-g, so if r=7% a stock with a 2% dividend yield implies a 5% growth rate whereas the company could increase its dividend so it yielded 3% but the growth rate would decline to 4% as it would not have that additional capital to reinvest (or it would have to issue new equity or increase debt ratios by additional borrowing, both of which would impact the valuation)

the model implies that a not paying a dollar of dividends today means that it has to pay (1+r)ⁿ dividends in year n. If a company elects to not pay a dividend, reinvests the dollar and earns less then management is destroying value. However if it has the ability to earn more than r then it should not pay a dividend and reinvest. This is how companies are managed - why the big tech stocks like Microsoft or Cisco pay dividends now rather than 20 years ago

 

[andrewkirk]

Just the plain old dividend discount model

In that case, you are not arguing that it doesn't matter whether a stock ever pays dividends, because there is no DDM value in that case. At most you are arguing that, if we accept the MM assumptions then, as long as a dividend stream is expected to eventually commence, it doesn't matter whether it commences now or at a future date, and the initial level of the dividends does not matter.

It appears then that we are not in disagreement, because that's what I've been saying all along, eg post 30:

all financial value of financial assets comes from discounted cash flow, either directly from the asset, or from using the ability to control that asset to benefit another financial asset. Those cash flows do not have to be in the next five years. But they must be planned to occur at some time.

 

[russ_watters]

It means that there will someday be a disaster that forces the value of the stocks on the market to zero (or close enough to not matter.)

Why? You said "will" there, but in your examples said "can"...and also singled out individual products, not the entire market. Those are big differences.

Your tulip market example implies you think all stock market values are 100% speculation and 0% tangable. That's true of beanie babies and tulips, but it is not true of stocks and as a result, there is no "will" about it.

The stock market's consistent long term growth is neither a fad nor a coincidence. It's a reflection of tangeable value.

 

[russ_watters]

I find it very easy to imagine more wrenching disruption than occurred in the 20th century. Almost every previous century had more wrenching disruption than the 20th. We just focus on the 20th more because it's more recent...

No, that's not a coincidence either. It is a natural product of civilization advancing. As a result, we have good reason to predict that this century will again be the most peaceful/stable in history.

 

[BWV]

In that case, you are not arguing that it doesn't matter whether a stock ever pays dividends, because there is no DDM value in that case. At most you are arguing that, if we accept the MM assumptions then, as long as a dividend stream is expected to eventually commence, it doesn't matter whether it commences now or at a future date, and the initial level of the dividends does not matter.

It appears then that we are not in disagreement, because that's what I've been saying all along, eg post 30:

Except that there are other buyers with other valuation metrics, if you are a control buyer - a strategic corporate buyer, activist hedge fund, private equity fund you could care less what the current or future dividend policy is and will value the asset off the total cash flows. The presence of these buyers will put a floor on stock values, which benefits individual investors

 

[andrewkirk]

control buyers could care less

Sorry for the tangential link BWV, but that phrase is a source of perpetual perplexity to me, and David Mitchell is always amusing to listen to, regardless of the topic.

we have good reason to predict that this century will again be the most peaceful/stable in history.

It will be great if you turn out to be right. I am less optimistic, but will rejoice if I turn out to be wrong (well, I would, but I don't expect, or hope, to be alive at the end of the century).

 

[Jeff Rosenbury]

Why? You said "will" there, but in your examples said "can"...and also singled out individual products, not the entire market Those are big differences.

Your tulip market example implies you think all stock market values are 100% speculation and 0% tangable. That's true of beanie babies and tulips, but it is not true of stocks and as a result, there is no "will" about it.

The stock market's consistent long term growth is neither a fad nor a coincidence. It's a reflection of tangeable value.

Given an infinitely long time, it will happen.

How the market reacts between now and then is the interesting bit.

Tulips and beanie babies have tangible value. They were just way overvalued by today's standards.

Value is determined by people. (At least until the robot apocalypse. ) Something is worth what people are willing to pay for it. And this amount changes through time. Should conditions change, value changes. It doesn't matter is the change is for the worse, like a government seizure, or for the better, like railroads replacing wagons. The world changes and value changes with it. There is more call for tulips today than there is for the more "tangible" wagon freight business of yesteryear.

 

[russ_watters]

Given an infinitely long time, it will happen.

Again, why? Because in 4 billion years the sun will swallow the Earth and vaporize the markets? You're being very vague, yet expressing a highly certain fatalistic view about this claim of yours.

Tulips and beanie babies have tangible value.

No they don't (at least not of any significance). Not that it is necessarily a reliable source, so I won't bother posting it, but a google tells me a Beanie Baby cost 38 cents to manufacture. It should be self-evident that a little scrap of fabric and some pvc pellet filler has essentially zero value outside of a fad.

And tulips, unless people eat them and I don't know about it, are flowers. Their value is aesthetic only. Not quite the same as a fad, but close.

Value is determined by people. (At least until the robot apocalypse. ) Something is worth what people are willing to pay for it. And this amount changes through time. Should conditions change, value changes. It doesn't matter is the change is for the worse, like a government seizure, or for the better, like railroads replacing wagons. The world changes and value changes with it. There is more call for tulips today than there is for the more "tangible" wagon freight business of yesteryear.

Again, an individual product or company going out of business does not imply that an entire stock market must eventually or can at any time go to zero. And saying that value is determined by people glosses over your claim, which is of a specific intrinsic value: zero. There are an infinite number of possible values that aren't zero and you need to explain why zero, specifically, is inevitable.

 

[andrewkirk]

And tulips, unless people eat them and I don't know about it, are flowers. Their value is aesthetic only.

As is a preference for a full stomach over an empty one.

As Jeff said, the (economic) value of something is what someone is prepared to pay for it. The cost of ingredients is irrelevant. I have never encountered an economic theory that says anything to the contrary.

 

[Jeff Rosenbury]

Again, why? Because in 4 billion years the sun will swallow the Earth and vaporize the markets? You're being very vague, yet expressing a highly certain fatalistic view about this claim of yours.

There are an infinite number of possible values that aren't zero and you need to explain why zero, specifically, is inevitable.

Because as long as the market has a higher value than the transaction cost, it will keep functioning. When I say zero, I mean less than the transaction cost.

Markets express chaotic behavior. As soon as someone figures an expression for the market value, it feeds back into the market and changes the expression. Thus market value has a large random component.

That means the value of the market will climb and fall. In an infinite time it will reach zero at some point. At that point the market ceases to function and goes away.

There are a large number of real events that can drive a market to the zero point. In addition to random psychological action, transaction costs can increase dramatically. Theft is usually cheaper than bargaining, and it's only social control that keeps transaction costs low. War, famine, disease, and lawlessness all raise transaction costs.

This is in addition to changing values due to changing desires, technological innovation, and changing physical needs.

Suppose for example someone expands the digital currency idea and forms an efficient digital stock market with tighter controls on graft. Transaction costs are driven down, so people stop using old markets like the NYSE. Without people using it, the NYSE can no longer afford maintenance, so it closes. This is just one example of millions that can drive a market to zero.

Humans have used many forms of social control/government historically. Now we are reaching an era when we start to question just what it is to be human. It is difficult (impossible) to foresee all eventualities, but it's safe to say tomorrow won't be like today.

So no, I can't be specific about how and why a market will reach zero.

 

[BWV]

control buyers could care less

Sorry for the tangential link BWV, but that phrase is a source of perpetual perplexity to me

I find that people capable of eating marmite can be easily perplexed ;)

Cool video though