Mr. Bogle Values the Market, 1928-2016

[SimpleGift]

In recent years, Mr. Bogle has written about and become well-known in press interviews for his simple method of estimating the fundamental expected return of the stock market:

• 

Using this simple formulation — and available financial data for the modern era — the chart below compares how Mr. Bogle would value the market (in blue) against how investors as a whole actually valued the market (in green) from 1928 to 2016.

• 
  Earnings growth, dividend yield and price data from Shiller

In general over the nine decades, the price return has not deviated too significantly from Mr. Bogle's fundamental return. In the 1940s and the 1970s, investors undervalued stocks relative to their fundamentals — then notably overvalued them in the 1990s. However, as if by gravity, the market price seems to always eventually revert back to its fundamental value. Thoughts?

[AlohaJoe]

Why is the Fundamental Return being compared to the Price Return? I would have expected it to be compared to the Total Return? Or is that just a typo?

[JoMoney]

It is practically a tautology. The return will always be that of combining the fundamental growth and the price multiple. We know the price multiple and the growth change over different periods... but despite knowing that they will converge and diverge, while always combining for the total return, we don't know how they will converge - if that means the fundamental growth will catch up to the price, or the price multiple will come up or down relative to the growth. It's an interesting way to look at it, but not especially helpful for making any decisions.

[lazyday]

the chart below compares how Mr. Bogle would value the market (in blue) against how investors as a whole actually valued the market (in green) from 1928 to 2016

So the 1 on the y axis is the scaled starting market price?

Could you explain what the blue fundamental line is charting?

Earnings growth + yield is the expected future return. I'm asking what your method is for converting that to a price.

[SimpleGift]

Earnings growth + yield is the expected future return. I'm asking what your method is for converting that to a price.

Same as for the price return which is scaled to 1.00 in 1927, the fundamental return (earnings growth rate + dividend yield) is also scaled to 1.00 in 1927, with annual changes to both index values charted thereafter — which gives a head-to-head matchup of changes to expected return and actual price return over the period.

[SimpleGift]

It's an interesting way to look at it, but not especially helpful for making any decisions.

True — though assuming one can accurately gauge the earnings growth rate at the current time, Mr. Bogle’s simple formula would have alerted an investor that stocks were significantly undervalued in the 1970s compared to their fundamentals, and likewise greatly overvalued in the late 1990s.

Not terribly useful to a buy-and-hold investor at the time — but potentially helpful as a gauge of market underpricing and overpricing.

[SimpleGift]

Why is the Fundamental Return being compared to the Price Return? I would have expected it to be compared to the Total Return? Or is that just a typo?

No, the Price Return is simply the market’s gauge of the value of the Fundamental Return at any given point in time. Sometimes the market overshoots in its estimation of value (irrational exuberance) and at other times undershoots (irrational pessimism). The chart in the OP is just to point out that it’s a self-correcting mechanism over long holding periods.

[lazyday]

Same as for the price return which is scaled to 1.00 in 1927, the fundamental return (earnings growth rate + dividend yield) is also scaled to 1.00 in 1927, with annual changes to both index values charted thereafter — which gives a head-to-head matchup of changes to expected return and actual price return over the period.

Are you saying that the y axis is not a price, but a rate of return? I'm not asking about the movement of the lines, but the axis itself, or where each line is at one moment of time.

[RadAudit]

Ah, the Gordon Equation!! Delightful little piece of information. See Bernstein's Four Pillars ...

If only I could remember to use it.

[abuss368]

Great article.

[abuss368]

Great article.

[SimpleGift]

Are you saying that the y axis is not a price, but a rate of return? I'm not asking about the movement of the lines, but the axis itself, or where each line is at one moment of time.

Perhaps I’m not understanding your question, but the y-axis is simply an index value for changes in both the price return and the fundamental return, both starting at 1.00 in 1927. The price return and the fundamental return are percentage changes each year, which are applied against the index value on the y-axis.

[lazyday]

OK, great! So if the market starts at 1000 and a year later is 1050, the green line moves from 1 to 1.05. To me that means that the y axis is a price.

My original question is how you are converting the fundamental return into a price. Does the blue line simply start with 1 (the market price of 1000 scaled down) and then never again look to the stock market price? Does it simply move each year by the amount of the current yield + expected growth rate?

Also, is the price return real or nominal?

[SimpleGift]

OK, great! So if the market starts at 1000 and a year later is 1050, the green line moves from 1 to 1.05. To me that means that the y axis is a price.

My original question is how you are converting the fundamental return into a price. Does the blue line simply start with 1 (the market price of 1000 scaled down) and then never again look to the stock market price? Does it simply move each year by the amount of the current yield + expected growth rate?

The y-axis is not a price, but a made up index value starting at 1.00 in 1927, for comparison purposes. BOTH the price return index value and the fundamental return index value move each year by the amount of their respective percentage changes. My apologies for not making this clearer in the OP. If it helps, here is an example of the type of chart it is:

• 

Both the price return and the earnings growth rate are nominal, as per Mr. Bogle's original formulation.

[lazyday]

Sorry for so many questions, I find the chart to be a strange one.

Is the earnings growth rate a constant, or do you change it over time? Is it simply the ex post historical nominal change in earnings per share, calculated recently?

Also, the chart in your last post shows total return, but your green line is a price return without dividends, right?

[SimpleGift]

Is the earnings growth rate a constant, or do you change it over time? Is it simply the ex post historical nominal change in earnings per share, calculated recently?

Also, the chart in your last post shows total return, but your green line is a price return without dividends, right?

No worries, I didn't expect the chart would be confusing, so didn't provide much explanation in the OP — my fault. The earnings growth rate is taken from the Shiller database (annual returns) and changes every year, just like the dividend yield. And, yes, the green line is price return only, without dividends, as this provides us with investors' assessment of market value over time.

[knpstr]

However, as if by gravity, the market price seems to always eventually revert back to its fundamental value. Thoughts?

In the short term the stock market is voting machine, in the long term it is a weighing machine. - Ben Graham

Prices ebb and flow around the true value (based upon fundamentals). The "trick" going forward is being able to correctly predict the future fundamental value.

[garlandwhizzer]

Great post, Simplegift. It makes the point that underlying fundamentals over the long haul have exerted gravity on market price euphoria and conversely given a lift to market price desperation from panic. Speculative return is powerful over short time spans but negligible over the long run. It also makes the point that investors should pay little or no attention to what is going on financially in the present and instead focus only on the long run, the opposite of what generally happens. Remember when watching CNBC to keep the mute button engaged and beware of slick marketing media by smart guys who promise to beat the market.

Bogle takes a modest view of future US stock returns, positive but well less than historical and bonds (somewhere around zero real) over the next decade. These predictions made in 2015 are based on fundamentals and so far have been wrong. The market has done much better than predicted in 2015, but Bogles piece doesn't attempt to predict the short term but instead focuses on the long term. If reversion to the fundamental curve occurs this time, it suggest that there may be a bit of trouble waiting for investors at some point down the road especially if market PEs revert to normal levels (which they never may do outside of a crash). I believe it wise to take his advice: don't make optimistic assumptions about future returns over the next decade. He has given other advice which should also be taken: "invest we must" and "stick to the plan whatever happens."

Garland Whizzer

[SimpleGift]

Speculative return is powerful over short time spans but negligible over the long run.

Just from curiosity, charted below are the annual deviations of the price index from the fundamental values in the OP. With today’s high market valuations, plus intense interest around the world in U.S. companies and their prospects, one wonders if there will ever be a future time when U.S. stocks again fall out of favor — like during the oil price shocks and high inflation years of the 1970s.

Barring a global catastrophe, it’s currently hard to envision another extended buying opportunity like back then.

• 

[Valuethinker]

Speculative return is powerful over short time spans but negligible over the long run.

Just from curiosity, charted below are the annual deviations of the price index from the fundamental values in the OP. With today’s high market valuations, plus intense interest around the world in U.S. companies and their prospects, one wonders if there will ever be a future time when U.S. stocks again fall out of favor — like during the oil price shocks and high inflation years of the 1970s.

Barring a global catastrophe, it’s currently hard to envision another extended buying opportunity like back then.

Alas mate, it's all too easy to imagine such a "buying opportunity". Which devastates the investment and savings plans of a generation.

Remember that that was only an extended buying opportunity if 1). you had capital or surplus income to invest 2). you didn't need the money to live on. The 1970s was when lots of people who had retired on fixed pensions, got wiped out. It was something of an accident, I believe, that US Social Security was indexed to CPI (rather than relying on ad hoc votes by the Congress).

No you just need to think about some of the crises that can impact the world economy, like a major interruption of oil supplies, or an all out war between NATO and Russian "volunteers" in the Baltic States, etc.

Or a major political crisis like Watergate.

History says the world throws up whoopsies.

Remember that the UK managed an 80-something per cent. fall in the stock market in real terms, with no revolution, no civil war, only mild civil unrest (1972-74). Of course in retrospect it was a once in a lifetime chance to buy stocks. But lots of people got wiped out.

media.morningstar.com/uk/AWS/Website/Crash_UK.pdf

says -73.81% but that's in nominal terms. Inflation was around 20% at that time from memory.

Nice listing in that reference of other bear markets, US and UK.

Or, for example, one of the historically famous American investment banks goes broke, and the result could be a catastrophic loss of confidence in the system . And a huge government bailout , coordinated globally. But what if next time someone gets up and screams "Britannia First" and the bailout doesn't happen?

Worse, what happens if an Austrian bank goes bust and there is a global convertibility of currencies, thus leading to a catastrophic fall in trade? .
Worse still, what happens then if an Austrian Corporal with delusions of grandeur is elected Chancellor of Germany, and seeks to unite all German speakers? As countries raise trade barriers, what happens if officers of a regional army in an East Asian country decide to stage a "Provocation" at a border crossing, and invade the other country in reprisal?

https://www.bloomberg.com/news/articles ... t-collapse

https://en.wikipedia.org/wiki/Mukden_Incident

What happens if an ex President's stockbroking firm goes bust, and this precipitates a financial crash in the US and the rest of the developed world, kicking off a decade or more of economic uncertainty?

https://en.wikipedia.org/wiki/Panic_of_1873

• 

What that says to me is that stock valuations, and in fact stock prices, are fractal, it's hard to argue there is reversion to the mean or that the deviations from the mean have a normal distribution.

The reality is that financial assets (and in particular stocks, although Mandelbrot's dataset was Egyptian cotton prices since the 1500s, daily data) don't have a central tendency-- they are in essence, fractal.

[Dirghatamas]

Simplegift
I just read Bogle's paper you linked while at lunch. I don't understand something basic about your equation. Why one piece of it is earning growth while the other is dividend yield, as opposed to earnings yield?

Here is the simple way I have intuitively always thought about fundamental (non speculative) returns:

Total fundamental returns = earnings yield% + earnings growth rate.

Thinking this way, treats dividends and buybacks equivalently. The reason I am pointing this out is that over the last hundred years, how earnings get paid back to shareholders has changed dramatically, with almost all of it paid back in dividends early on and nowadays majority in buybacks.

The opposite is true for other world markets where much/most of cash paid back is still in dividends. So, I think a model which simplifies this down to just earnings would be preferable (all of history, all country markets) to making the artificial distinction between dividends and buybacks (which after all is a distinction just about taxes).

[SimpleGift]

Here is the simple way I have intuitively always thought about fundamental (non speculative) returns:

Total fundamental returns = earnings yield% + earnings growth rate.

Thinking this way, treats dividends and buybacks equivalently. The reason I am pointing this out is that over the last hundred years, how earnings get paid back to shareholders has changed dramatically, with almost all of it paid back in dividends early on and nowadays majority in buybacks.

You’re opening a can of worms with your question, and I’m reluctant to get into the debate about dividends vs. share buybacks. Suffice it to say, as you suggest, there’s a strong economic argument that the two are functionally equivalent, since a shareholder can always sell stock in proportion to the buyback, and reinvest the money received in a diversified portfolio — which is the same as having received a dividend.

Mr. Bogle has always used “earnings growth rate + dividend yield” — which is what the OP examines historically — and I'll leave it there.

[Dirghatamas]

[
You’re opening a can of worms with your question, and I’m reluctant to get into the debate about dividends vs. share buybacks. Suffice it to say, as you suggest, there’s a strong economic argument that the two are functionally equivalent

Yes I am opening a can of worms but the linked paper you had from Bogle is literally titled "Occam's Razor Redux". In that Bogle spends the first long paragraph telling us that he wants to have the simplest possible model and go more complex only if needed (Occam's Razor). Let's see, if he uses an equation with earnings growth, he must have information about earnings. It seems simpler to have only one quantity: earnings and its rate of change rather than two quantities (one of which dividends is arbitrary) and the rate of change of one (earnings).

I would argue that Bogle's formula is more complex than it needs to be unless there is a fault in this line of thinking (simply thinking in terms of earnings and its rate of change). Simply saying Bogle wrote it this way, doesn't mean much. If you are going to post, you should be open to discussion which might help our collective understanding and (perhaps) lead to simpler models.

I don't see a problem in opening a can of worms if it leads to a simplification of our understanding.

[SimpleGift]

I don't see a problem in opening a can of worms if it leads to a simplification of our understanding.

OK. One argument I’ve seen is that, in Mr. Bogle’s simplified valuation formula, all the factors are stated per share. If earnings growth rates and dividend yields are calculated per share, then it doesn’t matter how many net shares are repurchased, since the effect is already included in these reported factors.

As mentioned, I’m reluctant to debate about alternative valuation methods in this thread, since we’ve already had many of those threads in the past (simply Google “Gordon equation” + bogleheads, if you're interested).

[siamond]

As JoMoney said, it's basically a tautology... when the 3rd factor (valuation) is added to the equation, which Jack typically does. For people moderately math-inclined, here is the explanation of such assertion.

The thing I never understood about Jack's writing in this respect is why he does the math in nominal terms instead of doing it in real terms like anybody else. Introducing the (highly unpredictable) inflation noise to the equation can only dilute the strength of the expected return math (which I personally checked while playing with such models at length). I suspect he essentially finds it simpler and more directly palatable. Sometimes a -perceived- simplification is just a bad idea imho.

[siamond]

Now the funny thing with this chart (or a chart showing total returns) is the remarkable straight line pattern that Prof. Siegel keeps referring to, heck he even has a so-called constant named about him!

It is really interesting to compare with other countries though. I compared to the UK and to Japan (because those are the only ones where I could find enough history), and assembled the following graph a while ago. It is also price-only, which is a tad unfortunate, but reliable historical dividends are hard to find...

One's imagination has to stretch to find a straight line in the UK returns. Now Japan... Yeah, maybe not. Personally I reached the clear conclusion that there is no such thing as a "Siegel's constant". Fundamental returns have no reason to be constant in average, in other words.

[Valuethinker]

Now the funny thing with this chart (or a chart showing total returns) is the remarkable straight line pattern that Prof. Siegel keeps referring to, heck he even has a so-called constant named about him!

It is really interesting to compare with other countries though. I compared to the UK and to Japan (because those are the only ones where I could find enough history), and assembled the following graph a while ago. It is also price-only, which is a tad unfortunate, but reliable historical dividends are hard to find...

One's imagination has to stretch to find a straight line in the UK returns.

http://www.poetrybyheart.org.uk/poems/t ... lish-road/

Before the Roman came to Rye or out to Severn strode,
The rolling English drunkard made the rolling English road.
A reeling road, a rolling road, that rambles round the shire,
And after him the parson ran, the sexton and the squire;
A merry road, a mazy road, and such as we did tread
The night we went to Birmingham by way of Beachy Head.

- G K Chesterton

Now Japan... Yeah, maybe not. Personally I reached the clear conclusion that there is no such thing as a "Siegel's constant". Fundamental returns have no reason to be constant in average, in other words.

Yes. In other words, stock returns don't follow nice patterns. Change the sample size (or the market) and you change the "constant".

Financial returns are more likely fractal than anything else. See Mandelbrot.

In his wonderful compedium of housing price data ("Safe as Houses: 8 centuries of housing prices") Neil Monnery actually shows that housing prices are *also* fractal, but he does not actually use the word.

[SimpleGift]

Fundamental returns have no reason to be constant in average, in other words.

Agreed, siamond. And if fundamental returns are not reasonably steady, one can’t expect stock returns to be steady also. The backstory to your nice chart upthread showing Japan’s price index history is that Japan’s corporate earnings fell into the ditch somewhere around 1990 (in green, chart below) and stock prices went in with them:

• 
  Source: MSCI Barra

This is certainly not the kind of earnings history one is used to seeing for the U.S. market. I don’t know enough about Japan’s corporate history to know why earnings collapsed so dramatically in the early 1990s, but it seems to show Mr. Bogle’s simple formula at work — stock price returns will eventually follow fundamental returns.

[garlandwhizzer]

Simplegift wrote:
ust from curiosity, charted below are the annual deviations of the price index from the fundamental values in the OP. With today’s high market valuations, plus intense interest around the world in U.S. companies and their prospects, one wonders if there will ever be a future time when U.S. stocks again fall out of favor — like during the oil price shocks and high inflation years of the 1970s.

Barring a global catastrophe, it’s currently hard to envision another extended buying opportunity like back then.

I agree with this point of view. There is a glut of savings collectively in the world although it seems to be concentrated in fewer and fewer hands. Still there's a lot of savings looking for safe and profitable investment opportunities these days. The US has taken center stage in that contest for a long time which is why our equity valuations are a bit stretched relative to the rest of the world. High quality fixed income offers essentially 0 -1% expected real returns over the next decade or so which further exacerbates the imbalance between massive savings on the one hand and fewer attractive places that offer low/modest risk and mid/high return. I believe there is a real possibility that we will not in the foreseeable future return to historical average PE levels in the US. PEs may have adjusted upward to a secular new normal in the US. Our economy is characterized by slow economic growth, high governmental debt and, for much of the population, high household debt as well, older population demographics, high savings concentrated among the rich investment class, diminishing productivity growth, stable government and economy, and low inflation. In such a scenario quality assets would be expected to be generously priced.

Historically, average US large cap PE between 1870 and 1982 were remarkably stable, no huge spikes up, but starting in 1982 when the S&P 500 had a PE of 7, a severe inflationary time, PEs have expanded dramatically. Even with that expansion in PE over more than 3 decades, for the entire time period 1870 - 2017, the median PE using "as reported earnings" was 15.66, a far cry from 25.69 where they are today. IMO we won't get back to that PE level in the US unless a major global catastrophic event occurs acutely. It may be mistaken to assume that mean reversion in US PE ratios will occur. Likewise it may be mistaken to assume that past historical returns will be future returns. Bill Bernstein argues that these changes may indicate a secular shift in wealthy stable DM countries which have a glut of savings chasing after limited opportunities for robust real positive gains without considerable risk.

This may explain why EM equity is bargain priced relative to the US now and for a few years. EMs are the opposite of us: fast economic growth, growing productivity, higher inflation, younger populations, lower levels of household debt, less investable savings, less stable governments, more corruption, and more political risk. The question is: is it attractive at these prices or a value trap? Opinions differ on that point. IMO it's a good idea to get some exposure to EM. If you can tolerate the volatility I believe you'll be rewarded in the long run.

Garland Whizzer

[Angst]

One of the typically propitious indicators I always keep an eye for is a new thread started by Simplegift. Thank you for your post!