Why is stock market return so consistent over time?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
strummer6969
Posts: 978
Joined: Tue Dec 15, 2020 10:59 am

Re: Why is stock market return so consistent over time?

Post by strummer6969 »

dboeger1 wrote: Thu Aug 11, 2022 3:43 am Note that demographics are not the only factor in determining growth. A certain amount of growth is by design in modern economics. Central banks and governments effectively manipulate currencies and markets to achieve desired effects. The US in recent years has targeted a 2% stable inflation rate, and seems to be willing to go to great lengths to maintain it, even in the face of a global pandemic, global supply shocks, etc. I don't want to get too political as I've gotten on the wrong side of the forum mods for dancing around this topic before, but I'll just say that to some extent, it doesn't matter who the winners and losers are of these policies, as long as the overall system remains stable and functional. Overall growth doesn't necessarily mean more or better opportunity for any given individual, just that someone somewhere is making more money on average. Expecting economic growth to mirror demographic trends 1-for-1 is a bit too narrow, in my opinion.
I believe the only two ways to have sustained real growth are to increase labor productivity and labor participation (which ties to demographics - birth rates, immigration). What other things do you think determine growth?

So tech can be a driver of productivity and using the monetary system to get more investment in innovation & tech can be a good thing, if they produce things that are worthwhile from a productivity perspective. I think that is the goal of modern economics & the innovation economy.
Independent George
Posts: 1592
Joined: Wed Feb 17, 2016 11:13 am
Location: Chicago, IL, USA

Re: Why is stock market return so consistent over time?

Post by Independent George »

mffl wrote: Wed Aug 10, 2022 11:57 pm
A couple time periods picked at random:
1871 to 1950: 6.61% real
1929 to 1960: 6.40% real
1960 to 2021: 6.61% real (I promise, I'm just picking these at random)\
First, those periods are not random - nor is the subject. In short, you're using stock market returns as a proxy for GDP growth over time since the industrial era, which has seen the exponential growth in standard of living for most of the world. Despite two world wars, colonialism, the depression, the cold war, pandemics, etc., this period accounts for more people around the world growing wealthier than at any other time in recorded history.

Second, this wealth is fundamentally not about resource extraction, but from the cumulative increase in knowledge and human capital that began before any of us was born. Melting sand into semiconductors is not limited by physical resources, but by ingenuity - and those semiconductors are then used to improve the world in ways that could not have been imagined a generation ago.

Will this go on forever? Not necessarily... but, call me an optimist, I think it will - or, if not forever, for far longer than anyone born today will live. If you watch the media, you'd think we're in a state of constant, brutal warfare and environmental catastrophe. If you read the academic literature, you'll know that we have far fewer wars than at any time in history, and despite the increased lethality of how we fight, in fact far fewer people die from it than in times past. We spent two years panicking over a pandemic that wouldn't have even been noticed a century ago. Space tourism exists - it's a thing that people do. And most importantly, all the knowledge that we've struggled to accumulate in human history continues to grow, and is now expanded upon by a far, far larger number of people across the globe than ever before. It's no longer the province of a handful of monks and wealthy individuals in the developed world - it's the combined effort of nearly the entire planet, and that knowledge is now transmitted, replicated, backed up, and expanded almost instantly.
moneywise3
Posts: 576
Joined: Mon Nov 22, 2010 5:54 pm

Re: Why is stock market return so consistent over time?

Post by moneywise3 »

Stock market is anything but consistent. Although consistent and non-emotional behavior on part of the investor can help them reap the benefits of it.
MarkRoulo
Posts: 1198
Joined: Mon Jun 22, 2015 10:25 am

Re: Why is stock market return so consistent over time?

Post by MarkRoulo »

mffl wrote: Fri Aug 12, 2022 8:19 am ... snip ...

Agreed. I'm not at all predicting that population growth will continue exponentially and unabated forever. You laid out many reasons why it's already started to slow many places and will likely peak in the coming decades and then total population may even start to decrease a bit!

But if we're purely looking backwards, population growth as an exponential factor is a pretty useful observation:

Image

The above is US population growth. World population growth looks similar.

If this trend is about to change, and it looks like it is, it's interesting to wonder how much of an effect that will have on long term economic trends.
If we are looking backwards, population growth as an exponent is pretty useful for the past few centuries. Prior to that, maybe not.

Guesstimates for the population of England for various years:
  • 1066: 1.7M
  • 1290: 4.8M (yay for the Medieval Warm Period)
  • 1400: 2.1M (boo, hiss for the Black Death)
  • 1450: 1.9M
  • 1600: 4.1M
Population growth really took off (in the west, at least) around the industrial revolution. Maybe it continues, but the TFR in wealthy countries is under 2 and unless Africa figures out a way to grow more food things are going to get dicey if population there continues to grow exponentially.

One example to illustrate (not prove) Africa's problem: Egypt is trying to feed ~100 million people using ~12,000 square miles of arable land. Egypt is poor.
User avatar
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Why is stock market return so consistent over time?

Post by SimpleGift »

mffl wrote: Fri Aug 12, 2022 8:19 am The above is US population growth. World population growth looks similar.

If this trend is about to change, and it looks like it is, it's interesting to wonder how much of an effect that will have on long term economic trends.
In fact, the annual growth rate of the world's population has already been declining for almost 60 years. It peaked sometime in the 1960s (chart below).
The question of the impact of this slowing population growth on global economic growth, on corporate earnings and on future stock returns is really unanswerable. So investing in stocks today requires a certain faith in the resiliency, the creativity and the ability of global corporate enterprise to provide positive real returns to shareholders. It's been a good bet for several centuries now, and is likely to still be a good bet in the future.
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Why is stock market return so consistent over time?

Post by vineviz »

Walkure wrote: Fri Aug 12, 2022 8:42 am
deanmoriarty wrote: Thu Aug 11, 2022 9:11 am
nisiprius wrote: Thu Aug 11, 2022 7:19 am
Over the time period 1965-1994, $10,000 would have grown to $35,900.89
Over the time period 1932-1961, $10,000 would have grownt to $204,381.23.

Would you describe that as "consistent?"
Your numbers are indeed thought-provoking, especially them being inflation adjusted.

What is the conclusion one should have from this? I would naively say: unless one plans to periodically and consistently invest over time AND significantly deleverage from equity to fixed income (a reasonable balance of nominal and inflation-protected) as they age, investing in the market for decades is a much wilder speculation than typically assumed?
I think part of the reason we perceive market returns as more stable/consistent than Nisi's chart would suggest is precisely because virtually everyone fits your naive "unless" case. I'd like to see a similar chart where, instead of growth of 10,000 we see growth of 1,000 a year for 30 years. I suspect that (1) the outcomes will have a much narrower dispersal (2) the very worst years in the lump sum chart will NOT be the worst DCA years as they bought more low for longer.
I don't find that to be the case. Using real (inflation-adjusted) values, investing $1k every year for 30 years produces the following result.

Minimum: $43,912.35
Maximum: $239,451.32

Image
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
dboeger1
Posts: 1411
Joined: Fri Jan 13, 2017 6:32 pm

Re: Why is stock market return so consistent over time?

Post by dboeger1 »

strummer6969 wrote: Fri Aug 12, 2022 9:07 am ]What other things do you think determine growth?
It depends on how you define growth. If you're talking about raw wealth, then sure, there's no getting around the laws of physics, you need more production to be wealthier. If you're talking in nominal currency terms, the value of a dollar isn't fixed. By some metrics, economies like Venezuela and Zimbabwe have gone through the roof. Of course, relative to the USD (world reserve currency), that's not the case, but my point was that the USD and other major currencies are manipulated too. Major economies try to walk the tightrope between monetary stability and growth.

There can be growth in nominal terms without there being a corresponding increase in real wealth. I forget the exact joke so I'll paraphrase, but a lawyer and a financial advisor walk past each other on a sidewalk. The adviser bumps into the lawyer, and the lawyer successfully sues the adviser for $1m. The adviser then charges the lawyer $1m to file the lawyer's now-more-complicated tax return. Neither one is wealthier, but look on the bright side, GDP just went up $2m! Now layer on top of that central bank policy, government spending, bond issuance, forex trading, etc., and it turns out the metrics don't really tell you much about real productivity beyond how they're strictly defined. Now, ideally, currency would have relatively stable value that reasonably approximates underlying wealth creation, but that doesn't always happen. How much money did governments around the world effectively pay people and companies so they wouldn't work out in public during the pandemic? As a result, official figures showed very modest economic declines, when in raw wealth terms, I have to believe the real losses to productivity were much more substantial, and we're seeing some of the after-effects now with inflation, low labor-force participation rates, etc.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Why is stock market return so consistent over time?

Post by skierincolorado »

Productivity, population growth and a bit of luck. Other markets are not quite so consistent
McQ
Posts: 1425
Joined: Fri Jun 18, 2021 12:21 am
Location: California

Re: Why is stock market return so consistent over time?

Post by McQ »

Hello OP: I would like to echo Nisiprius’ summary of where Jeremy Siegel went astray. The consistency that you thought you found is in part an artifact of Siegel’s data limitations. His sources got the pre-1871 US stock returns wrong.

Siegel also did not keep up with the updates that Dimson, Marsh & Staunton made in the two decades after publication of Triumph of the Optimists (search on “Credit Suisse yearbooks”). Triumph was published at the peak of the dotcom boom; in the years since, the authors have corrected survivorship bias by adding in the “losers” they had omitted. If you look country by country, you don’t see that consistency you thought you found in the US, itself again partly an artifact of Siegel’s outdated data sources.

It's still true that (US) stock returns become more consistent the longer the holding period, as this chart will show.

Image

But even at 50 years, with a dispersion in real returns between 4% and 9% , there’s plenty of room for the kinds of wealth disparity that Nisiprius demonstrates.

Last, you might enjoy two other recent forums, in which your issues are refracted through different lens: viewtopic.php?t=382854 and viewtopic.php?t=383542
You can take the academic out of the classroom by retirement, but you can't ever take the classroom out of his tone, style, and manner of approach.
ososnilknarf
Posts: 210
Joined: Mon Feb 18, 2019 6:08 pm

Re: Why is stock market return so consistent over time?

Post by ososnilknarf »

Independent George wrote: Fri Aug 12, 2022 9:31 am We spent two years panicking over a pandemic that wouldn't have even been noticed a century ago.
I get the gist of your post, but this statement I can't quite grasp what you mean. In the 1920's hospital workers in cities wouldn't have noticed that people were coming in critically ill at such a high rate that they ran out of space, ran out of body bags, ran out of space in the morgues? I just can't fathom how that could be possible. Even with today's incredible medical technology and vaccines and treatments 1 million people died in the US alone.
Am I missing something? Are you saying that people just didn't go to the hospital then and would have just died at home in bed and so no one would notice the trend?
User avatar
steve r
Posts: 1300
Joined: Mon Feb 13, 2012 7:34 pm
Location: Connecticut

Re: Why is stock market return so consistent over time?

Post by steve r »

Independent George wrote: Fri Aug 12, 2022 9:31 am
mffl wrote: Wed Aug 10, 2022 11:57 pm
A couple time periods picked at random:
1871 to 1950: 6.61% real
1929 to 1960: 6.40% real
1960 to 2021: 6.61% real (I promise, I'm just picking these at random)\
First, those periods are not random -
...
+1
The dates selected seemed a bit odd. Why count 1929 to 1950 and 1960 in multiple samples. If three periods are selected, why not 50 / 51 years each?
"Owning the stock market over the long term is a winner's game. Attempting to beat the market is a loser's game. ..Don't look for the needle in the haystack. Just buy the haystack." Jack Bogle
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

steve r wrote: Fri Aug 12, 2022 6:45 pm
Independent George wrote: Fri Aug 12, 2022 9:31 am
mffl wrote: Wed Aug 10, 2022 11:57 pm
A couple time periods picked at random:
1871 to 1950: 6.61% real
1929 to 1960: 6.40% real
1960 to 2021: 6.61% real (I promise, I'm just picking these at random)\
First, those periods are not random -
...
+1
The dates selected seemed a bit odd. Why count 1929 to 1950 and 1960 in multiple samples. If three periods are selected, why not 50 / 51 years each?
Completely fair point. I was just trying to recreate dart throwing I've done before with picking random time periods and being shocked at how often it's right at about 7%. Not trying to pick time periods that had any particular sense to them.

I think McQ's chart is the completed version of what I was trying to accomplish. They show that it's not right at 7% nearly as often as my dart throwing would suggest. Their chart and others have shown that the mean reversion (if it in fact exists) doesn't really start to manifest very consistently until 50 year holding periods. And even then the range is still 4% to 9%. But I still find myself fascinated by even that. I suppose we can all choose to be fascinated by different things. :) But it does seem like quite a few time periods on the 50 year graph are in the 6-7% range. Not as many as my initial post would suggest, but still very interesting. Actually, take that chart and bound it by 5% to 8% and see how many 20, 30, and 50 year holding periods fall in that range. Far from universal. Not trying to make a 100% stock argument here. 50 years is too long of a holding period for nearly any investor, and generally anyone with even close to a 50 year period will have an insignificant amount to invest. But still I find it amazing how many results fall in that band.
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

McQ wrote: Fri Aug 12, 2022 5:06 pm Hello OP: I would like to echo Nisiprius’ summary of where Jeremy Siegel went astray. The consistency that you thought you found is in part an artifact of Siegel’s data limitations. His sources got the pre-1871 US stock returns wrong.

Siegel also did not keep up with the updates that Dimson, Marsh & Staunton made in the two decades after publication of Triumph of the Optimists (search on “Credit Suisse yearbooks”). Triumph was published at the peak of the dotcom boom; in the years since, the authors have corrected survivorship bias by adding in the “losers” they had omitted. If you look country by country, you don’t see that consistency you thought you found in the US, itself again partly an artifact of Siegel’s outdated data sources.

It's still true that (US) stock returns become more consistent the longer the holding period, as this chart will show.

Image

But even at 50 years, with a dispersion in real returns between 4% and 9% , there’s plenty of room for the kinds of wealth disparity that Nisiprius demonstrates.

Last, you might enjoy two other recent forums, in which your issues are refracted through different lens: viewtopic.php?t=382854 and viewtopic.php?t=383542
Thank you! Prior to this discussion, I didn't even know about Siegel’s work let alone what was wrong with it, and now I know both. A very interesting discussion regardless, and thanks for your insight.

ETA: Also thank you for the chart, that is very enlightening.
User avatar
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Why is stock market return so consistent over time?

Post by SimpleGift »

mffl wrote: Fri Aug 12, 2022 11:22 pm But it does seem like quite a few time periods on the 50 year graph are in the 6-7% range. Not as many as my initial post would suggest, but still very interesting. Actually, take that chart and bound it by 5% to 8% and see how many 20, 30, and 50 year holding periods fall in that range....But still I find it amazing how many results fall in that band.
What you are seeing, I believe, is just the long-term impact on U.S. stock prices of the long-term fundamental growth of stocks' earnings-per-share. From 1935 to present, earnings-per-share growth has been pretty much in the 5%-7% range (chart below) — and stock prices have generally followed. This is Mr. Bogle's "fundamental return" of the market, as opposed to its "speculative return."
For long holding periods, it's the fundamental return that bears out over the shorter-term speculative gyrations.
Last edited by SimpleGift on Sat Aug 13, 2022 12:51 pm, edited 1 time in total.
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

SimpleGift wrote: Sat Aug 13, 2022 9:03 am
mffl wrote: Fri Aug 12, 2022 11:22 pm But it does seem like quite a few time periods on the 50 year graph are in the 6-7% range. Not as many as my initial post would suggest, but still very interesting. Actually, take that chart and bound it by 5% to 8% and see how many 20, 30, and 50 year holding periods fall in that range....But still I find it amazing how many results fall in that band.
What you are seeing, I believe, is just the long-term impact on U.S. stock prices of the long-term fundamental growth of stocks' earnings-per-share. From 1935 to present, real earnings-per-share growth has been pretty much in the 5%-7% range (chart below) — and stock prices have generally followed. This is Mr. Bogle's "fundamental return" of the market, as opposed to its "speculative return."
For long holding periods, it's the long-term fundamental return that bears out over the shorter-term speculative gyrations.
One fascinating aspect of this topic is how different charts make it look like it's a real thing, and others make it look like it's not a thing. Your chart makes it look very real. Nisiprius' chart makes it look like it's not. Both are valid representations of reality.
dbr
Posts: 46181
Joined: Sun Mar 04, 2007 8:50 am

Re: Why is stock market return so consistent over time?

Post by dbr »

mffl wrote: Sat Aug 13, 2022 10:23 am

One fascinating aspect of this topic is how different charts make it look like it's a real thing, and others make it look like it's not a thing. Your chart makes it look very real. Nisiprius' chart makes it look like it's not. Both are valid representations of reality.
So true, and it comes to what one thinks one means by "consistent." But even that chart above is over such a short period of time relative to the timing of all that noise that getting the appearance of an exponential fit could be deception more than reality. In my mind having to draw those border lines at such different rates of growth shows that up. Does anyone have an actual number for goodness of fit?

One can see here: https://en.wikipedia.org/wiki/Goodness_of_fit It has to do with the chances that what one thinks one sees is in fact not real. Related to that is the common question whether or not linear models are sufficient or indeed there is non-linearity = not consistent. With lots of noise you can't even test that. This sort of analysis discusses that point: https://socratic.org/questions/how-do-y ... ppropriate.

The other issue though, about which various comments have been made, is whether the model should be linear. If stock market trend is related to economic basics and those have an intrinsic character of log linear growth then the market should. But are there such factors and have they grown that way?

I sometimes think of a comparison to some phenomena such as why is the decay curve for radioactivity in a sample of radioactive material is exactly exponential. It turns out that arises from the two simple assumptions that all nuclei of a given isotope are exactly identical and that the probability of decay of any single nucleus in a unit of time has no time dependence. After that it is simple to derive an exponential decay function. Any variation in decay rate in any unit of time is simply a problem of large numbers (The Poisson distribution is derived) and the trend line has to be exact. Is there any such law that applies to the stock market and why? It might be the actual answer is that the market is consistent over time because it is.
User avatar
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Why is stock market return so consistent over time?

Post by SimpleGift »

Just as a historical footnote, Mr. Bogle's distinction between the fundamental and speculative return of the stock market did not originate with him, but rather grew from a seed planted in his mind by John Maynard Keynes, the legendary British economic theorist and author. Mr. Bogle acknowledged this debt to Keynes in a 2014 article:
John C. Bogle wrote:My ability to understand what would become my major field of study was no more than, shall we say, adequate. But of all the reading that I did in my field of concentration, it was Keynes's The General Theory of Employment, Interest, and Money, published in 1936, that has stayed at the forefront of my mind to this very day.

Although there is a lot of dense doctrine in that timeless book, I was particularly struck by chapter 12, "The State of Long-Term Expectation." There, Keynes made a critical distinction between the two broad reasons that explain the return of stocks. The first was what he called enterprise — "forecasting the prospective yield of an asset over its entire life." The second was speculation — "forecasting the psychology of the market."
It's interesting how such an idea, seemingly reflected from one mind into another mind, can then take root and grow — and thus has a life in human consciousness down through time.
User avatar
PicassoSparks
Posts: 417
Joined: Tue Apr 28, 2020 5:41 am

Re: Why is stock market return so consistent over time?

Post by PicassoSparks »

Saying the range is 5%-7% real as if that’s a small range is belied by the chart which shows 7% getting you to 100 while 5% gets you to under 20.

Even on a human-scale 40 year time horizon $1 at 5% compounds to about $7 and at 7% compounds to about $15.

Compounding breaks intuitions.
User avatar
Oicuryy
Posts: 1959
Joined: Thu Feb 22, 2007 9:29 pm

Re: Why is stock market return so consistent over time?

Post by Oicuryy »

This is Shiller's monthly real total return price data.
http://www.econ.yale.edu/~shiller/data.htm

Image

Excel added the exponential trend line. It is drawn with a log (base 2) scale on the y axis. The space between grid lines represents a doubling or halving.

Sometimes the actual value is twice the trend value and sometimes it's half. But it does not get much farther away than that. Is that close enough to be called consistent?

Ron
Money is fungible | Abbreviations and Acronyms
dbr
Posts: 46181
Joined: Sun Mar 04, 2007 8:50 am

Re: Why is stock market return so consistent over time?

Post by dbr »

Oicuryy wrote: Sat Aug 13, 2022 1:30 pm This is Shiller's monthly real total return price data.
http://www.econ.yale.edu/~shiller/data.htm



Excel added the exponential trend line. It is drawn with a log (base 2) scale on the y axis. The space between grid lines represents a doubling or halving.

Sometimes the actual value is twice the trend value and sometimes it's half. But it does not get much farther away than that. Is that close enough to be called consistent?

Ron
A next step often done in regression analysis, especially with a high R2, would be to check a residuals plot to be sure it is random. Probably what is not random in stock returns is the progression of secular bull and bear markets that you can even see in the data. Whether one wants to say that 10-20 year under-return followed by over-return on that timescale is a lack of consistency would need to be discussed.

I think everyone finds the remarkable fit to a fixed exponential function to be pretty convincing for the period shown.
User avatar
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Why is stock market return so consistent over time?

Post by SimpleGift »

mffl wrote: Sat Aug 13, 2022 10:23 am One fascinating aspect of this topic is how different charts make it look like it's a real thing, and others make it look like it's not a thing. Your chart makes it look very real. Nisiprius' chart makes it look like it's not. Both are valid representations of reality.
The difference in perspective seems to be a result of the difference in time scales we are looking at.

From the perspective of an individual stock investor — with an investing career of 20, 30, or 50 years — there has certainly been a wide dispersion in ending wealth amounts, depending on the particular period one is born into.

However from an academic perspective, we have return data on corporate enterprise extending over almost four centuries, going back to the 1600s and 1700s. Admittedly, the very early return data is quite weak — but even within this limited data set, stock returns on the scale of centuries have averaged a fairly consistent 5%-6% real per year.

For more, see this Forum discussion from several years ago: The Deep History of Stock & Bond Returns
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

PicassoSparks wrote: Sat Aug 13, 2022 12:28 pm Saying the range is 5%-7% real as if that’s a small range is belied by the chart which shows 7% getting you to 100 while 5% gets you to under 20.

Even on a human-scale 40 year time horizon $1 at 5% compounds to about $7 and at 7% compounds to about $15.

Compounding breaks intuitions.
Oh, completely agreed. But I think in a sense that simply describes how vastly different scenarios we might see if very long term trends ran very far outside of those bounds. Dark ages on one end, technological singularity on the other.
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

SimpleGift wrote: Sat Aug 13, 2022 2:26 pm
mffl wrote: Sat Aug 13, 2022 10:23 am One fascinating aspect of this topic is how different charts make it look like it's a real thing, and others make it look like it's not a thing. Your chart makes it look very real. Nisiprius' chart makes it look like it's not. Both are valid representations of reality.
The difference in perspective seems to be a result of the difference in time scales we are looking at.

From the perspective of an individual stock investor — with an investing career of 20, 30, or 50 years — there has certainly been a wide dispersion in ending wealth amounts, depending on the particular period one is born into.

However from an academic perspective, we have return data on corporate enterprise extending over almost four centuries, going back to the 1600s and 1700s. Admittedly, the very early return data is quite weak — but even within this limited data set, stock returns on the scale of centuries have averaged a fairly consistent 5%-6% real per year.

For more, see this Forum discussion from several years ago: The Deep History of Stock & Bond Returns
Right, this makes sense. So there really does seem to be consistency, but over perhaps 75 year supra human timescales rather than 20-30, and it's weak-ish even at 50. That perspective difference might be a large portion of why the various charts thrown around in this thread can all be true at the same time.

To some extent, though, I find that even odder, and that sort of begs my original question. Intuitively, I understand short term volatility, but if there were some long term consistency, it's over the 20-30 year horizon that I'd expect to see it, not the 75+ year horizon. Wouldn't you think reversion to the mean -- if it were to happen at all -- would have to happen on a timescale shorter than "human flight to internet"? Because you'd think that when the really big, once in a couple generation type things come along, they would break whatever the prevailing rate of inertia was, in one direction or the other. Yes, I suppose, the emergence of both flight and the internet were required to keep the compounded 7% thing going long term, but it just seems even odder that the fit is better the further you zoom out. Certainly that hasn't always been the case throughout human history.

So far my best grasp of this is that since the late 1800s, or perhaps even earlier to the 1600s, as you suggest, but perhaps NOT earlier than that, population growth has been pretty stubbornly exponential, especially in stable countries. That's certainly the cause of some of the growth, and it is in fact itself exponential and stable. Then if productivity is generally positive long term -- more erratic than population growth, but still bounded over long timescales -- then economic growth should be expected to be relatively consistent and exponential. And if the market is just the economy with a temper, it will -- eventually -- revert to the true value of the economy. Therefore, population growth is relatively consistent, even over shorter timescales. Economic growth less so (due to adding in the less stable productivity growth factor), and the market least of all, but if you zoom out far enough to smooth out the peaks and valleys, it's just a straight line too on a log chart.

So one theory might be that we're in for trouble if population growth continues to slow, as that contributes a large portion of the absolute economic growth, and nearly ALL of the consistency.

Another might be that population growth has already slowed, and the zoomed out rate of growth isn't problematic yet, not even with the 2018, 2020, and 2022 bear markets. In fact the tech era since the 80s has a higher return if anything. Perhaps technological productivity growth is supplanting population growth, and perhaps that's... ok? After all, Moore's law is a bit artificial and misunderstood, but still presents as a pretty consistent exponential curve itself. And perhaps if tech stopped propping us up, governments would start stressing out and heavily incentivizing larger families again, etc. And people would feel less wealthy, which might possibly tend to increase the birth rate. In other words, it might end up being self correcting if tech growth can't offset all the population growth.

Maybe. :)
User avatar
AnnetteLouisan
Posts: 7261
Joined: Sat Sep 18, 2021 10:16 pm
Location: New York, NY

Re: Why is stock market return so consistent over time?

Post by AnnetteLouisan »

Maybe Because they keep removing the companies that fail or get delisted in the year that occurs, or the year they fall out of the index, rather than counting their lack of earnings eternally in the chart? Plus they add new healthy ones?
Last edited by AnnetteLouisan on Sat Aug 13, 2022 5:28 pm, edited 3 times in total.
User avatar
AnnetteLouisan
Posts: 7261
Joined: Sat Sep 18, 2021 10:16 pm
Location: New York, NY

Re: Why is stock market return so consistent over time?

Post by AnnetteLouisan »

ososnilknarf wrote: Fri Aug 12, 2022 6:33 pm
Independent George wrote: Fri Aug 12, 2022 9:31 am We spent two years panicking over a pandemic that wouldn't have even been noticed a century ago.
I get the gist of your post, but this statement I can't quite grasp what you mean. In the 1920's hospital workers in cities wouldn't have noticed that people were coming in critically ill at such a high rate that they ran out of space, ran out of body bags, ran out of space in the morgues? I just can't fathom how that could be possible. Even with today's incredible medical technology and vaccines and treatments 1 million people died in the US alone.
Am I missing something? Are you saying that people just didn't go to the hospital then and would have just died at home in bed and so no one would notice the trend?
They noticed in 1918.
User avatar
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Why is stock market return so consistent over time?

Post by SimpleGift »

mffl wrote: Sat Aug 13, 2022 5:09 pm So one theory might be that we're in for trouble if population growth continues to slow, as that contributes a large portion of the absolute economic growth, and nearly ALL of the consistency.

Perhaps technological productivity growth is supplanting population growth...
It's nearly certain that the growth rate of the global population will continue to slow over the 21st century. This has been the forecast by the United Nations Population Division for some time (their median variant scenario), and their population forecasts going back to the 1950s have proven surprisingly accurate.

Whether global productivity growth will take up the slack in the future is an open question. The Brookings Institute has published a series of solid articles addressing these issues, if interested: Growth in a Time of Change

All in the "food for thought" category. Keep in mind that pure economics discussions (non-investing related) are against this Forum's rules though.
Last edited by SimpleGift on Sat Aug 13, 2022 6:05 pm, edited 1 time in total.
User avatar
nisiprius
Advisory Board
Posts: 52215
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Why is stock market return so consistent over time?

Post by nisiprius »

AnnetteLouisan wrote: Sat Aug 13, 2022 5:21 pm
ososnilknarf wrote: Fri Aug 12, 2022 6:33 pm
Independent George wrote: Fri Aug 12, 2022 9:31 am We spent two years panicking over a pandemic that wouldn't have even been noticed a century ago.
I get the gist of your post, but this statement I can't quite grasp what you mean. In the 1920's hospital workers in cities wouldn't have noticed that people were coming in critically ill at such a high rate that they ran out of space, ran out of body bags, ran out of space in the morgues? I just can't fathom how that could be possible. Even with today's incredible medical technology and vaccines and treatments 1 million people died in the US alone.
Am I missing something? Are you saying that people just didn't go to the hospital then and would have just died at home in bed and so no one would notice the trend?
They noticed in 1918.
There was a degree of wartime secrecy about the flu because governments did not want each other to know how the flu was affecting their military capability. One reason it was called the "Spanish flu" was because Spain was neutral and wasn't censoring all mention of it.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
Oicuryy
Posts: 1959
Joined: Thu Feb 22, 2007 9:29 pm

Re: Why is stock market return so consistent over time?

Post by Oicuryy »

dbr wrote: Sat Aug 13, 2022 2:14 pm A next step often done in regression analysis, especially with a high R2, would be to check a residuals plot to be sure it is random.
Are residuals the distances between the price line and the trend line? (Statistics is not my strong suit.) If so, here they are plotted in units of grid lines (powers of 2).

Image

mffl, reversion to trend occurs occurred in less than 75 years. The zero line in this graph is the trend line. Look at the distance between dots that touch the zero line. To my eye the longest stretch was a little over 200 months – about 20 years.

There were also times when the price went from way above trend to way below trend. And vice versa. To have gotten trend growth you would have needed to get in and out at points that were the same distance from the zero line.

Ron
Money is fungible | Abbreviations and Acronyms
bh1
Posts: 326
Joined: Thu Dec 30, 2021 1:49 pm

Re: Why is stock market return so consistent over time?

Post by bh1 »

snackdog wrote: Thu Aug 11, 2022 2:01 pm 7% is the natural return. If it were much higher too many people would pile in. If it were lower, nobody would bother.
The "market" (meaning people) make an estimation of expected return on investment multiplied by some risk factor. High returns and/or low risk compared to the market average means more people buy, driving the price up, while low returns and/or high risk have their stock price drop until market average. It so happens that the market average is about 7%. This is just a rewording of the above.

I think corporations themselves target their return on investment to be about 7%, as much higher is risky and lower is comparatively unprofitable. Thus, the 7% is a self-fulfilling prophesy. The return has been 7% because everyone is targeting historically acceptable returns or 7%.
dbr
Posts: 46181
Joined: Sun Mar 04, 2007 8:50 am

Re: Why is stock market return so consistent over time?

Post by dbr »

Oicuryy wrote: Sat Aug 13, 2022 6:52 pm
dbr wrote: Sat Aug 13, 2022 2:14 pm A next step often done in regression analysis, especially with a high R2, would be to check a residuals plot to be sure it is random.
Are residuals the distances between the price line and the trend line? (Statistics is not my strong suit.) If so, here they are plotted in units of grid lines (powers of 2).

Image

mffl, reversion to trend occurs occurred in less than 75 years. The zero line in this graph is the trend line. Look at the distance between dots that touch the zero line. To my eye the longest stretch was a little over 200 months – about 20 years.

There were also times when the price went from way above trend to way below trend. And vice versa. To have gotten trend growth you would have needed to get in and out at points that were the same distance from the zero line.

Ron
So those residuals seem to show us a periodic oscillation around a trend rather than random noise. Whether we call that consistent or not probably depends on what timescale we are interested in. We already know that in fact the stock market is plagued by secular bull and bear periods and the time scale, while not at all certain might be your 20 years or sometimes less to complete one side of the oscillation or 40 years or less to complete a complete cycle. It doesn't seem predictable enough to start running around saying this secular bull market is ending and that secular bear market is starting.

What those residuals don't show is a right side up or right side down "u" shape that would suggest there should be curvature added to the shape, which means mathematically that in the regression of log market against time one should add a quadratic in time or some other addition to a linear fit.

So the description for the time considered is that log market is linear in time with a superimposed oscillation that sort of runs every 30 or 40 years for a full cycle. I think there is too much irregularity in the departure to yap too much about periodicity while the dominating log-linear thing at 95% R2 is amazing.

Someone who works with this sort of financial stuff professionally could enlighten us I am sure.
Valuethinker
Posts: 49033
Joined: Fri May 11, 2007 11:07 am

Re: Why is stock market return so consistent over time?

Post by Valuethinker »

MarkRoulo wrote: Fri Aug 12, 2022 10:00 am
Population growth really took off (in the west, at least) around the industrial revolution. Maybe it continues, but the TFR in wealthy countries is under 2 and unless Africa figures out a way to grow more food things are going to get dicey if population there continues to grow exponentially.
Or. They import more food from places like the Ukraine. Which was the largest external supplier of food to North Africa and the Middle East.

The problem in Africa is poverty, not shortage of food per se. The same is true of India, say.

"grow exponentially" is a problematic term. Do we mean to the power of 2? Or to the power of -0.5?

In the long run, if your TFR is below 2.1 (probably higher than that for Least Developed Countries) then your population should, in theory, start to decline. That's already true for roughly half the population of the planet (? Europe, North America, most of East Asia, most of Latin America, some of the Middle East), and looks likely to be true in the 2nd half of the 21st century for most of it.
One example to illustrate (not prove) Africa's problem: Egypt is trying to feed ~100 million people using ~12,000 square miles of arable land. Egypt is poor.
Egypt actually has some of the best agricultural land in the world. However the Aswan High Dam has screwed up the hydrology. And precious water gets spent growing cash crops like cotton - which is notoriously water hungry - and not food.

Nonetheless Egypt has been a net food importer for a long time. But then England has been a net food importer since at least the 1700s (& maybe before). Again, it's not the inability to produce food domestically which is the main problem, it's the lack of income.

Africa is in trouble. And sub Saharan Africa is the one place where TFR has not fallen fast (enough). Rising temperatures and lack of rainfall plus demographics could lead to disaster & mass migration.

But it's wrong to think the world can't feed Africa. It can. Remembering that the world's fastest rising health problem is obesity - -even in places like China and India.
Valuethinker
Posts: 49033
Joined: Fri May 11, 2007 11:07 am

Re: Why is stock market return so consistent over time?

Post by Valuethinker »

AnnetteLouisan wrote: Sat Aug 13, 2022 5:21 pm
ososnilknarf wrote: Fri Aug 12, 2022 6:33 pm
Independent George wrote: Fri Aug 12, 2022 9:31 am We spent two years panicking over a pandemic that wouldn't have even been noticed a century ago.
I get the gist of your post, but this statement I can't quite grasp what you mean. In the 1920's hospital workers in cities wouldn't have noticed that people were coming in critically ill at such a high rate that they ran out of space, ran out of body bags, ran out of space in the morgues? I just can't fathom how that could be possible. Even with today's incredible medical technology and vaccines and treatments 1 million people died in the US alone.
Am I missing something? Are you saying that people just didn't go to the hospital then and would have just died at home in bed and so no one would notice the trend?
They noticed in 1918.
They certainly did. The canonical example of St Louis v Philadelphia.

The St Louis governing Council slammed down stiff restrictions. Cinemas and bathhouses were closed (in those days, many people would have bathed at a bath house, lacking a bath at home). The Philadelphia Mayor insisted on holding a Victory Bond parade - estimates suggesting something like 30k people died as a result of that parade?

1-2% of the population of the world died*. But in some places, such as Alaskan tribes, it was so bad that it is considered impolite to this day to speak of it-- a bit like wandering into a Jewish house and casually mentioning the Holocaust. Also in Africa and India we read of whole villages being wiped out - but the colonial masters weren't bothered with keeping that accurate records.

What is interesting is the historical memory blank of it. Look on most peoples' family tree, and you will find "oh that was Dad's aunt Sadie. She died in the Flu". But we stopped talking about it. Maybe the horrors of World War One, the elation of the 1920s Jazz era, the privations of the Great Depression - maybe that just overwrote the memories.

* there is a wide range of estimates, I am giving you the one I vaguely recall.
Last edited by Valuethinker on Sun Aug 14, 2022 4:34 pm, edited 1 time in total.
Valuethinker
Posts: 49033
Joined: Fri May 11, 2007 11:07 am

Re: Why is stock market return so consistent over time?

Post by Valuethinker »

mffl wrote: Sat Aug 13, 2022 5:09 pm
SimpleGift wrote: Sat Aug 13, 2022 2:26 pm
mffl wrote: Sat Aug 13, 2022 10:23 am One fascinating aspect of this topic is how different charts make it look like it's a real thing, and others make it look like it's not a thing. Your chart makes it look very real. Nisiprius' chart makes it look like it's not. Both are valid representations of reality.
The difference in perspective seems to be a result of the difference in time scales we are looking at.

From the perspective of an individual stock investor — with an investing career of 20, 30, or 50 years — there has certainly been a wide dispersion in ending wealth amounts, depending on the particular period one is born into.

However from an academic perspective, we have return data on corporate enterprise extending over almost four centuries, going back to the 1600s and 1700s. Admittedly, the very early return data is quite weak — but even within this limited data set, stock returns on the scale of centuries have averaged a fairly consistent 5%-6% real per year.

For more, see this Forum discussion from several years ago: The Deep History of Stock & Bond Returns
Right, this makes sense. So there really does seem to be consistency, but over perhaps 75 year supra human timescales rather than 20-30, and it's weak-ish even at 50. That perspective difference might be a large portion of why the various charts thrown around in this thread can all be true at the same time.

To some extent, though, I find that even odder, and that sort of begs my original question. Intuitively, I understand short term volatility, but if there were some long term consistency, it's over the 20-30 year horizon that I'd expect to see it, not the 75+ year horizon. Wouldn't you think reversion to the mean -- if it were to happen at all -- would have to happen on a timescale shorter than "human flight to internet"? Because you'd think that when the really big, once in a couple generation type things come along, they would break whatever the prevailing rate of inertia was, in one direction or the other. Yes, I suppose, the emergence of both flight and the internet were required to keep the compounded 7% thing going long term, but it just seems even odder that the fit is better the further you zoom out. Certainly that hasn't always been the case throughout human history.

So far my best grasp of this is that since the late 1800s, or perhaps even earlier to the 1600s, as you suggest, but perhaps NOT earlier than that, population growth has been pretty stubbornly exponential, especially in stable countries. That's certainly the cause of some of the growth, and it is in fact itself exponential and stable. Then if productivity is generally positive long term -- more erratic than population growth, but still bounded over long timescales -- then economic growth should be expected to be relatively consistent and exponential. And if the market is just the economy with a temper, it will -- eventually -- revert to the true value of the economy. Therefore, population growth is relatively consistent, even over shorter timescales.
Population growth is anything but consistent? It rises quite rapidly as life expectancies lengthen, and then slows dramatically. France in the 19th Century does not have fast population growth. Victorian Britain does, and then it does not. Every other industrialising country follows the same pattern into the 20th Century.
Economic growth less so (due to adding in the less stable productivity growth factor), and the market least of all, but if you zoom out far enough to smooth out the peaks and valleys, it's just a straight line too on a log chart.

So one theory might be that we're in for trouble if population growth continues to slow, as that contributes a large portion of the absolute economic growth, and nearly ALL of the consistency.
1. Population growth is not consistent. 2. As long as GDP per capita grows, actual population growth is not a big issue.

Just raising the rest of the planet up to the standard of living of Costa Rica will keep us busy this century. Costa Rica is a place where the poor live in shacks - that have running water & electric lights & TVs. And they have free universal healthcare, and free education up to age 18. And the life expectancy and literacy of a developed country.
Another might be that population growth has already slowed, and the zoomed out rate of growth isn't problematic yet, not even with the 2018, 2020, and 2022 bear markets. In fact the tech era since the 80s has a higher return if anything. Perhaps technological productivity growth is supplanting population growth, and perhaps that's... ok? After all, Moore's law is a bit artificial and misunderstood, but still presents as a pretty consistent exponential curve itself.
Productivity growth surged in the lead up to 2000, but has been quite flat since. Technology is not obviously causing productivity growth.

The real key with population is "the Demographic Dividend". See Japan post WW2. Or Ireland in the 1980s & 90s. China since about 1990. Once Total Fertility Ratios drop, you have this situation over the subsequent 20 years where the workforce keeps growing, and the number of dependents falls. Often associated with large scale migration of women into the formal workforce. The Dependency ratio (dependents per worker) falls.

The next phase is where Italy is now. Or Japan. Where the number of old people rises dramatically and the Dependency Ratio goes into reverse. Japan has struggled with this & they seem to be far better at providing for their old than many western countries. So that's a warning.
And perhaps if tech stopped propping us up, governments would start stressing out and heavily incentivizing larger families again, etc. And people would feel less wealthy, which might possibly tend to increase the birth rate. In other words, it might end up being self correcting if tech growth can't offset all the population growth.

Maybe. :)
A couple of points there:

1. although US social policy provides little support to families to have more children, AFAIK, natalist policies have been widely tried since the 1930s in other countries. Ceausescu's Romania banning abortion & artificial means of birth control, etc. Massive incentives for more children (Nazi Germany first, but many countries in modern post war Europe).

Broadly these have been unsuccessful. European nations, the most advanced practitioners, have TFRs generally below 2.1 ie below replacement. Former communist Eastern Europe and Southern Europe particularly low.

2. What we do know is:

- the more religious people are the more children they tend to have - Chassidic Jews, Amish, Mormons etc. So if there is another big religious revival in the Western World then you might see a birthrate surge.

- extensions in life expectancy tend to bring about a fall in birth rates. But there's a lag, as more of the over 35s are alive but so are their children. So there is a distinct "hump" in the population for c 2 generations whilst this works its way through. Since 1900, world life expectancies have x2 with less developed countries increasing by the most.

- related to that is when do people marry, and do women have a place in higher education and in the formal workforce? Garry Becker and the Chicago economics school seem to be right. Children are an enormous economic opportunity cost in terms of lost income & careers etc. And women bear the disproportionate share of that. To the extent that women can delay having children, they will have fewer... or none at all.

Countries that have experienced (mild) revivals in TFR, like France and Scandinavia, seem to be:
- ones where there is a lot of state support for families. For example universal, cheap childcare
- housing prices are not exiguous - young couples can afford apartments or homes
- men take on a greater share of childcare and housework. For example paternity leave is both offered, and taken

None of the above is true of places with really low birthrates like Spain, Japan, Italy.

Urban India now has TFRs of around 1.8-1.9 vs 3.1 for the country as a whole (down from over 6 in 1960). That is quite striking.

(the biology of deferral of having children is pretty stark. I encourage anyone to research the odds. Remember that when Mrs Blair (British PM Tony Blair's wife, and an accomplished trial lawyer, a QC) had her 4th child in her 40s, she was already clearly one of the women in this world who was quite fertile -- so she was not representative of the odds of any randomly selected 40-something British woman having a child).

3. the Baby Boom is an odd one. It seems related to a deferral of child-bearing due to the exigencies of the Great Depression and the Second World War. It really was only a significant factor in the overseas Anglosphere - Australia, Canada, USA. It was much later and more muted in Europe (including Britain).

The Baby Boom is one of the very few (only?) recorded rises in TFR since the 1850s in a developed country*. The hypothesis is that people did so much better in the late 1940s & 1950s than their parents had, or than that they had expected to growing up in the 1930s, that having children seemed a smaller cost. And the returning soldiers and sailors tended to marry very quickly upon return. And women left, voluntarily or not, the majority of the "male" roles they had occupied during the war years when men were in uniform.

https://academic.oup.com/book/404 Sara Harper Demographics: A Very Short Introduction is an excellent and readable introduction to these issues.

I don't think demographics is much help in determining the course of stock prices.

* Russia after the 1918-21 Civil War. But Russia was not a developed country in that time - industrialisation under Stalin came later.
Valuethinker
Posts: 49033
Joined: Fri May 11, 2007 11:07 am

Re: Why is stock market return so consistent over time?

Post by Valuethinker »

SimpleGift wrote: Fri Aug 12, 2022 10:04 am
mffl wrote: Fri Aug 12, 2022 8:19 am The above is US population growth. World population growth looks similar.

If this trend is about to change, and it looks like it is, it's interesting to wonder how much of an effect that will have on long term economic trends.
In fact, the annual growth rate of the world's population has already been declining for almost 60 years. It peaked sometime in the 1960s (chart below).
The question of the impact of this slowing population growth on global economic growth, on corporate earnings and on future stock returns is really unanswerable. So investing in stocks today requires a certain faith in the resiliency, the creativity and the ability of global corporate enterprise to provide positive real returns to shareholders. It's been a good bet for several centuries now, and is likely to still be a good bet in the future.
Many people do not realise this. About falling growth rates of population.

Thank you for reproducing that chart.
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

Valuethinker wrote: Sun Aug 14, 2022 4:35 pm
SimpleGift wrote: Fri Aug 12, 2022 10:04 am
mffl wrote: Fri Aug 12, 2022 8:19 am The above is US population growth. World population growth looks similar.

If this trend is about to change, and it looks like it is, it's interesting to wonder how much of an effect that will have on long term economic trends.
In fact, the annual growth rate of the world's population has already been declining for almost 60 years. It peaked sometime in the 1960s (chart below).
The question of the impact of this slowing population growth on global economic growth, on corporate earnings and on future stock returns is really unanswerable. So investing in stocks today requires a certain faith in the resiliency, the creativity and the ability of global corporate enterprise to provide positive real returns to shareholders. It's been a good bet for several centuries now, and is likely to still be a good bet in the future.
Many people do not realise this. About falling growth rates of population.

Thank you for reproducing that chart.
Thank you. This and your other post are very insightful.
CraigTester
Posts: 1488
Joined: Wed Aug 08, 2018 6:34 am

Re: Why is stock market return so consistent over time?

Post by CraigTester »

SimpleGift wrote: Sat Aug 13, 2022 9:03 am
mffl wrote: Fri Aug 12, 2022 11:22 pm But it does seem like quite a few time periods on the 50 year graph are in the 6-7% range. Not as many as my initial post would suggest, but still very interesting. Actually, take that chart and bound it by 5% to 8% and see how many 20, 30, and 50 year holding periods fall in that range....But still I find it amazing how many results fall in that band.
What you are seeing, I believe, is just the long-term impact on U.S. stock prices of the long-term fundamental growth of stocks' earnings-per-share. From 1935 to present, earnings-per-share growth has been pretty much in the 5%-7% range (chart below) — and stock prices have generally followed. This is Mr. Bogle's "fundamental return" of the market, as opposed to its "speculative return."
For long holding periods, it's the fundamental return that bears out over the shorter-term speculative gyrations.
Hopefully I’m not deviating too far from the threads topic, but your above chart brings up a CAPE related thought.

The core criticisms of CAPE I’ve seen on this forum and elsewhere tend to center on how the measuring of Earnings has changed over time. (E.g. accounting changes, tax changes, etc, etc).

And therefore CAPE (or PEn) has been rendered a useless metric whose current elevated value is simply a shortcoming of these changes.

However your chart does not seem to reflect this argument. Rather the long term Earnings trend remains remarkably consistent. (E.g. not trending lower to reflect measurement technique changes)

So perhaps we can’t explain away CAPE’s extended elevated values, by criticizing the metric itself?

Which perhaps leads us to just chalk up today’s elevated valuations to what Bogle termed “speculative return”, after all…?

Helpful to understand if true….
Valuethinker
Posts: 49033
Joined: Fri May 11, 2007 11:07 am

Re: Why is stock market return so consistent over time?

Post by Valuethinker »

AnnetteLouisan wrote: Sat Aug 13, 2022 5:17 pm Maybe Because they keep removing the companies that fail or get delisted in the year that occurs, or the year they fall out of the index, rather than counting their lack of earnings eternally in the chart? Plus they add new healthy ones?
Broadly that's true. Companies drop out of the index, new, successful companies are added.

"The creative and destructive forces of capitalism". I think that's a quote from Karl Marx, who wrote quite majestically about everything capitalism had accomplished in transforming the world*. This really became the basis of Joseph Schumpeter's economics. Whereas John Maynard Keynes (his contemporary) created the modern field of macroeconomics, working from the top down, Schumpeter was instrumental in the study of the "black box" - what actually goes on inside firms & industries to make these revolutions possible-- to enable the change we experience in the world.



* Marx wrote about industrialisation and had a quite neat schema about the transition from slavery to serfdom to industrialisation to liberal democracy to socialism. What modern historical work has added is a much better understanding of how this happened in the colonies, and generally in places outside of Europe. The centrality of the Slave Trade to the Early Modern World and the prosperity of 1600s & 1700s England which enabled the Industrial Revolution to take place. The role of colonialism in supporting industrialisation. What happened in the period 1750 (earlier, actually, it had already begun in the late 1600s) to 1900 was unprecedented in human history - and largely built upon the backs of slave labour (and some quite cute exploitation of foreign lands).
dbr
Posts: 46181
Joined: Sun Mar 04, 2007 8:50 am

Re: Why is stock market return so consistent over time?

Post by dbr »

I still haven't seen an explanation for just exactly why a log-linear fit with a superimposed bull-bear cycle should be expected in the US stock market over the last hundred years plus. One always has a sneaking suspicion that the mathematical result is somehow an artifact, by chance, over such a small domain (range of time) that linearity is just an approximation that does not last, etc.

To list reasons why there should be growth does not list reasons why that growth in an exponential model should be so consistent. Note the growth is not consistent in any other model. It is rapidly increasing in a linear model and would be decreasing in a log-log model.

We have already discussed whether or not we think there is a lot of noise around the very consistent log-linear trend.
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

dbr wrote: Mon Aug 15, 2022 1:00 pm I still haven't seen an explanation for just exactly why a log-linear fit with a superimposed bull-bear cycle should be expected in the US stock market over the last hundred years plus. One always has a sneaking suspicion that the mathematical result is somehow an artifact, by chance, over such a small domain (range of time) that linearity is just an approximation that does not last, etc.

To list reasons why there should be growth does not list reasons why that growth in an exponential model should be so consistent. Note the growth is not consistent in any other model. It is rapidly increasing in a linear model and would be decreasing in a log-log model.

We have already discussed whether or not we think there is a lot of noise around the very consistent log-linear trend.
Thank you, this is a very good rephrasing of my original question. Lots of very interesting discussion as to whether there is actually reversion to trend, how much noise there is, the expected timescale for trend reversion, whether there's consistency in the periodicity of the trend reversion, the extent to which population growth or productivity contributes to the growth, etc.

But... why? :)

Excellent commentary by Valuethinker and others, but I still can't shake that the strongest/most stable corrolary to the exponential market trend over time (over the last 200-400 years, at least) is population growth as the comparatively stable exponential baseline, plus other factors on top (productivity and...?). If it's not that, then what? Again, it's not enough to explain how growth occurs at all, it's the stability of the long term trend that's so staggering.

Perhaps our inflation adjustments, which are obviously limited/flawed to some degree by nature, are creating a feedback loop in the data, where inflation adjusted growth over time is 7% because the inflation adjustments assume something about growth rates, perhaps as a second or third order effect. But I'm not aware of the mechanism in BLS methodology that would cause this, and it's awfully consistent over time to not be a first order effect, isn't it?
User avatar
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Why is stock market return so consistent over time?

Post by SimpleGift »

mffl wrote: Mon Aug 15, 2022 3:51 pm ...but I still can't shake that the strongest/most stable corrolary to the exponential market trend over time (over the last 200-400 years, at least) is population growth as the comparatively stable exponential baseline, plus other factors on top (productivity and...?). If it's not that, then what?
One perspective to consider is that corporate enterprise has been a successful social phenomenon in different societies over several centuries, and has consistently grown shareholders' real returns over long holding periods at a rate of about 5% per year. In other words, the growth over long periods is intrinsic to the corporate model.

One interesting data point is the Bazacle company in France, the earliest and most long-lived documented shareholding company in the world, which operated over six centuries from 1372-1946. What were its shareholders' average real returns over the lifetime of the company? Yes, they averaged 5% real per year. See this Forum thread for more on this: One Company, Six Centuries of Stock Returns, 1372-1946
Last edited by SimpleGift on Mon Aug 15, 2022 4:30 pm, edited 1 time in total.
User avatar
PicassoSparks
Posts: 417
Joined: Tue Apr 28, 2020 5:41 am

Re: Why is stock market return so consistent over time?

Post by PicassoSparks »

We’ve spent the last 200 years drawing down the planet’s capital reserves of hydrocarbon. This cheap and dense source of energy and material has enabled a tremendous acceleration in economic productivity thanks to mechanization, fertilization, mass production, and transportation. There’s more people but there’s also more production per person.
Walkure
Posts: 1023
Joined: Tue Apr 11, 2017 9:59 pm

Re: Why is stock market return so consistent over time?

Post by Walkure »

vineviz wrote: Fri Aug 12, 2022 10:20 am
Walkure wrote: Fri Aug 12, 2022 8:42 am
deanmoriarty wrote: Thu Aug 11, 2022 9:11 am
nisiprius wrote: Thu Aug 11, 2022 7:19 am
Over the time period 1965-1994, $10,000 would have grown to $35,900.89
Over the time period 1932-1961, $10,000 would have grownt to $204,381.23.

Would you describe that as "consistent?"
Your numbers are indeed thought-provoking, especially them being inflation adjusted.

What is the conclusion one should have from this? I would naively say: unless one plans to periodically and consistently invest over time AND significantly deleverage from equity to fixed income (a reasonable balance of nominal and inflation-protected) as they age, investing in the market for decades is a much wilder speculation than typically assumed?
I think part of the reason we perceive market returns as more stable/consistent than Nisi's chart would suggest is precisely because virtually everyone fits your naive "unless" case. I'd like to see a similar chart where, instead of growth of 10,000 we see growth of 1,000 a year for 30 years. I suspect that (1) the outcomes will have a much narrower dispersal (2) the very worst years in the lump sum chart will NOT be the worst DCA years as they bought more low for longer.
I don't find that to be the case. Using real (inflation-adjusted) values, investing $1k every year for 30 years produces the following result.

Minimum: $43,912.35
Maximum: $239,451.32

Image
Very interesting. It's a slightly smaller range of outcomes (545% max/min vs 569% for the lump sum) but not as much I would have expected. Out of curiousity, is the $1k contribution being adjusted for inflation each year, or is it chipping in a nominal $30k over the period and just inflation-adjusting the returns?
Also, it shifts the worst starting year forward by over a decade, with the minimum balance coming from the period starting in 1952, compared to the lump sum minimum in 1965 (although '55/'56 were not much better).
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Why is stock market return so consistent over time?

Post by vineviz »

Walkure wrote: Mon Aug 15, 2022 4:45 pm Out of curiousity, is the $1k contribution being adjusted for inflation each year....
Yes. The contribution is inflation-adjusted.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

SimpleGift wrote: Mon Aug 15, 2022 4:27 pm
mffl wrote: Mon Aug 15, 2022 3:51 pm ...but I still can't shake that the strongest/most stable corrolary to the exponential market trend over time (over the last 200-400 years, at least) is population growth as the comparatively stable exponential baseline, plus other factors on top (productivity and...?). If it's not that, then what?
One perspective to consider is that corporate enterprise has been a successful social phenomenon in different societies over several centuries, and has consistently grown shareholders' real returns over long holding periods at a rate of about 5% per year. In other words, the growth over long periods is intrinsic to the corporate model.

One interesting data point is the Bazacle company in France, the earliest and most long-lived documented shareholding company in the world, which operated over six centuries from 1372-1946. What were its shareholders' average real returns over the lifetime of the company? Yes, they averaged 5% real per year. See this Forum thread for more on this: One Company, Six Centuries of Stock Returns, 1372-1946
Very interesting. For those who didn't read the article but are wondering, they didn't go out of business in 1946, they were nationalized along with every other utility on France at that time.
Topic Author
mffl
Posts: 235
Joined: Wed Dec 06, 2017 6:25 pm

Re: Why is stock market return so consistent over time?

Post by mffl »

PicassoSparks wrote: Mon Aug 15, 2022 4:30 pm We’ve spent the last 200 years drawing down the planet’s capital reserves of hydrocarbon. This cheap and dense source of energy and material has enabled a tremendous acceleration in economic productivity thanks to mechanization, fertilization, mass production, and transportation. There’s more people but there’s also more production per person.
Excellent point, though that's an explanation for a large jump in production, or an increase in the rate of return prior to the industrial revolution, but it still doesn't explain the flatness of the very long term trend since that time. I can see how hydrocarbons = giant jump in production, but not how hydrocarbons = 7%pa.
Dusn
Posts: 217
Joined: Thu Feb 15, 2018 8:59 am

Re: Why is stock market return so consistent over time?

Post by Dusn »

What’s more important for the long term growth of our current stock market index funds - the world population or the population in developed nations (in particular the US)?

As long as developed nations allow immigrants to move and work in their countries, I don’t see why a worldwide decline in population will necessarily affect them?
User avatar
PicassoSparks
Posts: 417
Joined: Tue Apr 28, 2020 5:41 am

Re: Why is stock market return so consistent over time?

Post by PicassoSparks »

Counter-intuitively, stock market returns are somewhat uncorrelated with economic growth. Because economic growth means a dilution of shares (new firms won’t be represented in your initial investment).
https://www.msci.com/documents/10199/a1 ... adb948f578
supply-side models tie a country’s stock returns to its GDP growth, but they do not suggest a perfect match between the two variables. Instead, they view real GDP growth as a cap on long-run stock returns, as other factors dilute GDP before it reaches shareholders
Post Reply