The index fund 6.5% myth

[oddballstocks]

 

I believe there is a tremendous body of evidence that clearly states that aggregate market PER-SHARE, PRE-TAX, REAL, TOTAL returns (at the country level) are not *at all* correlated with long term GDP growth.  I think they are certainly correlated (in certain ways, maybe with a lead or lag) with changes in GDP growth from expectations.

 

 

There is a good book that has data that provides a very compelling case for inverse correlation between GDP growth and stock market returns.

 

Triumph of the Optimists: 101 Years of Global Investment Returns

 

Vinod

 

So this would mean that over long periods stock returns are always negative right?  If countries continue go grow (GDP growth) then returns should be negative over the long haul?  If you extrapolate out a bit it would mean that equity investing is a losing game, as the world grows we're trying to swim against the tide.  Yet this doesn't seem to be the market experience.

 

Since we have numbers the US has grown as an economy and our market has been biased upwards.  So how does this jive with the research?

 

Dont take it too literally. It just means it would be a a bad idea to expect higher equity returns just because it has a higher growth rate.

 

The data if you care to look in the book, shows that countries that have higher growth rate are priced higher. So equity returns tended to disappoint investors who naively investing in high growth countries. Does this remind you of value investing?

 

Part of the explanation is also that to fund additional growth you need to make additional investments via debt and stock market issues which dilute existing shareholders.

 

Vinod

 

I guess I was thinking about this with a longer time window than most.  I was thinking if you looked at 100/200/300 years of history what would you see?  Someone mentioned short time periods, and in that vein I agree that growth in GDP doesn't correlate to returns.  I'd like to see the data from 1906 to 2006 for example, what's the relationship between those two?  My guess is that GDP growth and market growth are fairly tightly connected.  All of the pessimism and optimism cancel out eventually.

[yadayada]

 

I believe there is a tremendous body of evidence that clearly states that aggregate market PER-SHARE, PRE-TAX, REAL, TOTAL returns (at the country level) are not *at all* correlated with long term GDP growth.  I think they are certainly correlated (in certain ways, maybe with a lead or lag) with changes in GDP growth from expectations.

 

 

There is a good book that has data that provides a very compelling case for inverse correlation between GDP growth and stock market returns.

 

Triumph of the Optimists: 101 Years of Global Investment Returns

 

Vinod

 

So this would mean that over long periods stock returns are always negative right?  If countries continue go grow (GDP growth) then returns should be negative over the long haul?  If you extrapolate out a bit it would mean that equity investing is a losing game, as the world grows we're trying to swim against the tide.  Yet this doesn't seem to be the market experience.

 

Since we have numbers the US has grown as an economy and our market has been biased upwards.  So how does this jive with the research?

 

Dont take it too literally. It just means it would be a a bad idea to expect higher equity returns just because it has a higher growth rate.

 

The data if you care to look in the book, shows that countries that have higher growth rate are priced higher. So equity returns tended to disappoint investors who naively investing in high growth countries. Does this remind you of value investing?

 

Part of the explanation is also that to fund additional growth you need to make additional investments via debt and stock market issues which dilute existing shareholders.

 

Vinod

 

I guess I was thinking about this with a longer time window than most.  I was thinking if you looked at 100/200/300 years of history what would you see?  Someone mentioned short time periods, and in that vein I agree that growth in GDP doesn't correlate to returns.  I'd like to see the data from 1906 to 2006 for example, what's the relationship between those two?  My guess is that GDP growth and market growth are fairly tightly connected.  All of the pessimism and optimism cancel out eventually.

http://dqydj.net/sp-500-return-calculator/

http://www.measuringworth.com/growth/index.php

 

1.7% increase in real value of s&P 500. So S&P 500 went up less then GDP went up. But most of the return came from reinvesting earnings. I think this is heavily skewed.

 

But if you look at 1955 to 2013, value of S&P 500 increased 2.7%, vs GDP of 3.1%. I think the increase here is because more products were produced by very large companies.

 

For example if you look at 1975 to 2013, S&P real value increases over 4% , and real GDP only increased 2.9%. This also explains the higher profit margins? More consolidated industries, and more consumer products produced by fewer large corporations.

 

When the walmarts of this world wipe out small companies not on stock exchanges, or too small for S&P500, the take from large corporations from GDP goes up.

 

http://answers.google.com/answers/threadview?id=523867

 

That site actually says there werent even 500 companies in S&P 500 untill 1957! That also explains a lot I guess. A lot of these companies must have been private then or something?

[rkbabang]

When the walmarts of this world wipe out small companies not on stock exchanges, or too small for S&P500, the take from large corporations from GDP goes up.

 

 

Exactly. The error people are making above with defining what the returns "must be" is mistaking the S&P500 (or some other index) for the economy as a whole.  When you are investing in any subset of the economy (every index is a subset, in fact all the publicly traded companies put together are only a subset of the economy)  then your returns can and will be different from the economy as a whole in one direction or the other.

 

[yadayada]

Yeah another thing is international earnings growth and globalization. And the fact that there is a huge rise in middle class in developing countries. Looking at KO walmart etc, sometimes more then 50% of profits come from international sales. That is why before 1950, it is pretty much flat, vs 3% growth , and then after 1975 the S&P grows faster then GDP, with massive rise of emerging countries.

 

So you have to wonder, will this continue? Given that a large part of consolidation is behind us? You would expect the S&P to stay flat, or maybe grow only a little from now on? Since you will not see that explosive international growth relative to current earnings.

[rkbabang]

Yeah another thing is international earnings growth and globalization. And the fact that there is a huge rise in middle class in developing countries. Looking at KO walmart etc, sometimes more then 50% of profits come from international sales. That is why before 1950, it is pretty much flat, vs 3% growth , and then after 1975 the S&P grows faster then GDP, with massive rise of emerging countries.

 

So you have to wonder, will this continue? Given that a large part of consolidation is behind us? You would expect the S&P to stay flat, or maybe grow only a little from now on? Since you will not see that explosive international growth relative to current earnings.

 

If you look at the standard of living in the United States, then take a look at the other 6.9 billion people, I think there is a lot of potential growth left.

 

[vinod1]

 

I believe there is a tremendous body of evidence that clearly states that aggregate market PER-SHARE, PRE-TAX, REAL, TOTAL returns (at the country level) are not *at all* correlated with long term GDP growth.  I think they are certainly correlated (in certain ways, maybe with a lead or lag) with changes in GDP growth from expectations.

 

 

There is a good book that has data that provides a very compelling case for inverse correlation between GDP growth and stock market returns.

 

Triumph of the Optimists: 101 Years of Global Investment Returns

 

Vinod

 

So this would mean that over long periods stock returns are always negative right?  If countries continue go grow (GDP growth) then returns should be negative over the long haul?  If you extrapolate out a bit it would mean that equity investing is a losing game, as the world grows we're trying to swim against the tide.  Yet this doesn't seem to be the market experience.

 

Since we have numbers the US has grown as an economy and our market has been biased upwards.  So how does this jive with the research?

 

Dont take it too literally. It just means it would be a a bad idea to expect higher equity returns just because it has a higher growth rate.

 

The data if you care to look in the book, shows that countries that have higher growth rate are priced higher. So equity returns tended to disappoint investors who naively investing in high growth countries. Does this remind you of value investing?

 

Part of the explanation is also that to fund additional growth you need to make additional investments via debt and stock market issues which dilute existing shareholders.

 

Vinod

 

I guess I was thinking about this with a longer time window than most.  I was thinking if you looked at 100/200/300 years of history what would you see?  Someone mentioned short time periods, and in that vein I agree that growth in GDP doesn't correlate to returns.  I'd like to see the data from 1906 to 2006 for example, what's the relationship between those two?  My guess is that GDP growth and market growth are fairly tightly connected.  All of the pessimism and optimism cancel out eventually.

 

The book I mentioned looked at the 101 year data from 1900 to 2000 and came to the conclusion I mentioned. The authors periodically publish an updated version of the data for a report they do for institutional investors so if you are interested you might find the data you are looking for there - but the conclusion does not change.

 

Their conclusions were something along the lines that as long as there is positive GDP growth, the GDP growth rate itself has weakly negative correlation to stock market returns.

 

Vinod

[vinod1]

Relevant quote from the book:

Figure 11-6 also shows the total real return from equities in each country (the green bars).

Visually, real returns seem unrelated to GDP growth, and statistically, the correlation is -0.27

for 1900–2000 and -0.03 for 1951–2000. These findings, while based on much longer periods

than earlier research, are not new. Siegel (1998) found that from 1970–97, the correlation

between stock returns and GDP growth was -0.32 for seventeen developed countries, and

-0.03 for eighteen emerging markets.

 

Vinod

[yadayada]

So isnt it a better idea to invest in emerging markets? They will likely have products that could do well internationally? And that will relatively provide much larger growth compared to their earnings now vs S&P 500 now?

[krazeenyc]

So isnt it a better idea to invest in emerging markets? They will likely have products that could do well internationally? And that will relatively provide much larger growth compared to their earnings now vs S&P 500 now?

 

Do you know of any well established consumer brand companies that primarily sell into these fast growing markets that are trading at reasonable valuations?

[Pretium]

yadayada, I too hesitate when recommending index funds to my friends.

 

Yes, index funds have given investors a fair return for the past 30 years. But the historical data itself is meaningless without context.

 

Howard Marks always stresses: Are we extrapolating or analyzing? How will index funds perform moving forward in a different world?

 

Horizon Kinetics does a great job analyzing the environment of the past 30 years in the attached document (page 6 in particular).

 

In my line of business, it is so easy to recommend index funds because it automatically lowers the cost to the client. It's therefore easier to justify your own consulting fees. Problem is, this is happening across the board without analysis of the underlying indices. For institutions, it still makes sense to invest in index funds regardless of whether you think you can find better active managers. Why? Nobody will blame you for picking an index fund. You're with the crowd.

 

I'm not exactly sure how impactful index funds and ETFs have been on individual security prices. There seems to be a lot of disagreement in the industry. Certainly, Horizon Kinetics thinks they've distorted some markets. But at the end of the day, I always worry when something becomes a default option without a second thought. An investing decision made without regard to price worries me. Doesn't matter if S&P 500 P/E is 22x or 12x, just buy the index.

 

Here's a quote from "The Outsiders" about Henry Singleton:

 

Just two years before Singleton died, Leon Cooperman asked him about the many Fortune 500 companies that had begun large stock repurchases, a practice that was now considered a prudent action. Singleton, who pioneered that strategy, replied "If everyone’s doing them, there must be something wrong with them."

Q4_2014_Commentary_FINAL.pdf

[yadayada]

So isnt it a better idea to invest in emerging markets? They will likely have products that could do well internationally? And that will relatively provide much larger growth compared to their earnings now vs S&P 500 now?

 

Do you know of any well established consumer brand companies that primarily sell into these fast growing markets that are trading at reasonable valuations?

My guess is the korean stock exchange?

 

My guess is , the US had a huge advantage in this because of their huge cultural influence in the world.

[benhacker]

Regarding negative vs. non-correlated.

 

I think Vinod put in the quote (or a quote) on this.  But to be clear, those who claim that the "returns" (again, loaded, you must define what you mean) are uncorrelated or "slightly" negatively correlated (like I do) really technically mean "the excess, real, per-share, total return, pre-tax is" slightly negatively correlated.  I think excess return is key, because in general, the return for equities in all markets in all growth scenarios is positive... so it's a question of 3% real vs 7% real returns.

 

I would agree that to take this negatively correlated argument to the extent of say a 20% / year economy or a -5% / year economy would be problematic, but we also have virtually zero datasets of such countries in the real world during modern times with trading equity markets.

 

Nate, to your comment:

 

<i>"I guess I was thinking about this with a longer time window than most.  I was thinking if you looked at 100/200/300 years of history what would you see?  Someone mentioned short time periods, and in that vein I agree that growth in GDP doesn't correlate to returns.  I'd like to see the data from 1906 to 2006 for example, what's the relationship between those two?  My guess is that GDP growth and market growth are fairly tightly connected.  All of the pessimism and optimism cancel out eventually."</i>

 

I hear you, but my comment was specifically meant to address this... when I say "LONG TERM" I mean.... LONG TERM like what you are saying.

 

Again, your statement makes so much sense, but is wrong once you add "per-share" to your definition.  Yes, growing countries have growing markets... and growing markets by size for equities.  But not all equity growth benefits investors from Time 0 to Time 100... a bunch of it comes through external capital, dilution, etc.

 

Credit Suisse had a great data book on this that covers all markets for equities around the world since 1880 or something... quite eye opening.  It's non-intuitive at first, but it's really quite a powerful idea.

 

As Yadayada says, the outgrowth of his (as I stated, incorrect) believe that growth of GDP relates to higher returns... is just to buy all high growth emerging market economies.  That may or may not be a good strategy from today's prices / values... but it's prima facie not a good strategy for the growth itself... because all market participants (AND COMPANIES) already think they know where the growth will be; and they are investing / competing accordingly.

 

Probably just repeating myself at this stage, so hopefully that is clear.

[philly value]

I think the evidence being cited does not support the statement implied, namely that "GDP growth does not impact [or negatively impacts] stock returns"

 

If you are talking about the entire market, then it is blindingly obvious that higher GDP growth, all else equal, equates to higher stock returns. But as I posted earlier, in addition to GDP growth, there is also (a) dividend yield, basically, what rate of capital reinvestment is required to generate the GDP growth, (b) corporate profit margins and tax rates, and © valuation multiples. Those are, mathematically, the four key forces influencing pre-tax, whole-market stock returns.

 

So I guess the argument is really that there is a certain relationship between GDP growth and the remaining three variables that is causing GDP growth and stock returns to not have the same positive correlation one would see if they held fixed all other factors.

 

But to me, that doesn't seem to be important in understanding the point of this topic. The point was, will the returns of the whole market [presumably, the whole U.S. market] be lower in the future than over the past 50 years. If one believes that profit margins, tax rates, and valuation multiples are unlikely to be more favorable in the future than today, and that GDP growth is likely to be lower as well, then this conclusion follows mathematically regardless of the studies being cited.

 

 

[KCLarkin]

If you are talking about the entire market, then it is blindingly obvious that higher GDP growth, all else equal, equates to higher stock returns.

 

You have two companies with the following stable profiles:

 

China Inc

Growth = 10%

ROE = 5%

P/B = 1

 

America Inc

Growth = 5%

ROE = 10%

P/B = 1

 

Which one will have higher total stock returns?

[philly value]

If you are talking about the entire market, then it is blindingly obvious that higher GDP growth, all else equal, equates to higher stock returns.

 

You have two companies with the following stable profiles:

 

China Inc

Growth = 10%

ROE = 5%

P/B = 1

 

America Inc

Growth = 5%

ROE = 10%

P/B = 1

 

Which one will have higher total stock returns?

 

"All else equal" means holding margins fixed.

[KCLarkin]

If you are talking about the entire market, then it is blindingly obvious that higher GDP growth, all else equal, equates to higher stock returns.

 

You have two companies with the following stable profiles:

 

China Inc

Growth = 10%

ROE = 5%

P/B = 1

 

America Inc

Growth = 5%

ROE = 10%

P/B = 1

 

Which one will have higher total stock returns?

 

"All else equal" means holding margins fixed.

 

Okay, which of these will produce better returns?

 

China Inc

Growth = 10%

ROE = 5%

P/B = 1

 

Canada Inc

Growth = 5%

ROE = 5%

P/B = 1

[Uccmal]

To the original point I have had a shift in sentiment.  I think index funds are good for non experienced investors.  On the other hand we watched the markets do nothing for ten years from 2000-2010.  You still get your dividends and dividend increases but until the tax situation is made more favourable to dividends in the US this is a lower return than historically.  I dont know what people will do with the advice I may give.  So, with markets more fairly valued versus earnings and earnings potential I am more inclined to advise them to Pay Down Their Mortgage, and all other debts.  Most people will be way further ahead doing it the old fashioned way. 

 

If they have more incomes than outgoes, and no debt, then they dont need my help. 

[philly value]

@KCLarkin: I assume by growth you mean revenue growth. If so, the answer is that both will provide the same stock return. However, unless you assume constant outside investment, the China stock is not sustainable. Forgetting ROE because leverage will complicate things, if you have a stock w/ 5% ROA and 100% reinvestment, in order to get 10% revenue growth, you need to assume that operating margins continually decline. I did this as a six year calculation, and in year 6, operating margin had been reduced to 4% from 5%. So it does not fit the requirement that margins remain the same.

 

But you raise a really interesting point that may be applicable to the situation in China and/or other economies growing quickly. If you assume that the firm is able to raise capital equal to 5% of its beginning assets each year, you can have 10% revenue growth while holding margins fixed, and with return for the original investors remains at 5% per year.

 

However, I think in the U.S, and in general looking at the aggregate market instead of one company, this is less important of a consideration over long time periods. If you are looking at the world stock market, then movements of capital between equities will not increase the total stock, so the question is if you anticipate capital flowing from other assets into stocks.

[yadayada]

it seems to me ROE and ROIC will be better in emerging economies? There is less competition, demand outstrips supply more, less regulation and often the tax situation is better. And all those companies in their early stages have less bureacracy. Finally when looking at some of those emerging market companies, they often do business in other asian countries as well. And if there is only a small difference in ROE or ROIC, but your paying less then half, it takes a long time to catch up for the higher return company.

[cobafdek]

Relevant quote from the book:

Figure 11-6 also shows the total real return from equities in each country (the green bars).

Visually, real returns seem unrelated to GDP growth, and statistically, the correlation is -0.27

for 1900–2000 and -0.03 for 1951–2000. These findings, while based on much longer periods

than earlier research, are not new. Siegel (1998) found that from 1970–97, the correlation

between stock returns and GDP growth was -0.32 for seventeen developed countries, and

-0.03 for eighteen emerging markets.

 

I'm not sure how seriously we should accept these researchers's interpretations.  Negative correlations in the 0.03 to 0.32 ranges would be laughed at by physicists, chemists, and engineers.  These are extremely small and weak correlations.  You've all seen random scatterplot diagrams in which some statistician draws a line through the mist and says there's a correlation.  Generally speaking, correlations of at least 0.70 - 0.80 are considered strong in the social sciences.  Moreover, the P-value in Ritter's paper was 0.16, i.e., not statistically significant.

 

If I remember my probability/statistics classes correctly, the statisticians square the correlation coefficients ® to come up with an R-square number, which represents how much of the variance in one variable explains the variance in the other.  So GDP growth "explains"  only 0.09% - 10% of the subsequent stock returns. 

 

Another illustration of the gulf between the physical science and economic/social science.  I think the most that can be made of this is that economic growth, by itself, is not meaningfully associated with equity returns.

[KCLarkin]

Relevant quote from the book:

Figure 11-6 also shows the total real return from equities in each country (the green bars).

Visually, real returns seem unrelated to GDP growth, and statistically, the correlation is -0.27

for 1900–2000 and -0.03 for 1951–2000. These findings, while based on much longer periods

than earlier research, are not new. Siegel (1998) found that from 1970–97, the correlation

between stock returns and GDP growth was -0.32 for seventeen developed countries, and

-0.03 for eighteen emerging markets.

 

I'm not sure how seriously we should accept these researchers's interpretations.  Negative correlations in the 0.03 to 0.32 ranges would be laughed at by physicists, chemists, and engineers.  These are extremely small and weak correlations.  You've all seen random scatterplot diagrams in which some statistician draws a line through the mist and says there's a correlation.  Generally speaking, correlations of at least 0.70 - 0.80 are considered strong in the social sciences.  Moreover, the P-value in Ritter's paper was 0.16, i.e., not statistically significant.

 

If I remember my probability/statistics classes correctly, the statisticians square the correlation coefficients ® to come up with an R-square number, which represents how much of the variance in one variable explains the variance in the other.  So GDP growth "explains"  only 0.09% - 10% of the subsequent stock returns. 

 

Another illustration of the gulf between the physical science and economic/social science.  I think the most that can be made of this is that economic growth is not meaningfully associated with equity returns.

 

In the quote, he is pretty clear that it seems unrelated for all the reasons you mentioned. The potential country level  dataset is too small to have any definitive conclusions,

[cobafdek]

Relevant quote from the book:

Figure 11-6 also shows the total real return from equities in each country (the green bars).

Visually, real returns seem unrelated to GDP growth, and statistically, the correlation is -0.27

for 1900–2000 and -0.03 for 1951–2000. These findings, while based on much longer periods

than earlier research, are not new. Siegel (1998) found that from 1970–97, the correlation

between stock returns and GDP growth was -0.32 for seventeen developed countries, and

-0.03 for eighteen emerging markets.

 

I'm not sure how seriously we should accept these researchers's interpretations.  Negative correlations in the 0.03 to 0.32 ranges would be laughed at by physicists, chemists, and engineers.  These are extremely small and weak correlations.  You've all seen random scatterplot diagrams in which some statistician draws a line through the mist and says there's a correlation.  Generally speaking, correlations of at least 0.70 - 0.80 are considered strong in the social sciences.  Moreover, the P-value in Ritter's paper was 0.16, i.e., not statistically significant.

 

If I remember my probability/statistics classes correctly, the statisticians square the correlation coefficients ® to come up with an R-square number, which represents how much of the variance in one variable explains the variance in the other.  So GDP growth "explains"  only 0.09% - 10% of the subsequent stock returns. 

 

Another illustration of the gulf between the physical science and economic/social science.  I think the most that can be made of this is that economic growth is not meaningfully associated with equity returns.

 

In the quote, he is pretty clear that it seems unrelated for all the reasons you mentioned. The potential country level  dataset is too small to have any definitive conclusions,

 

Yup, yet they seem to like to publicize statements like this (from the abstract of Ritter's paper):  ". . . the cross-country correlation of real stock returns and per-capita GDP growth over 1900-2002 is negative."  It's these bald-faced assertions that get the headlines, and sticks in people's memories.

[yadayada]

why not use common sense? you can clearly see accelerated growth from 1950 to 2000, where at first it lags, and then it grows faster , right around when emerging markets start to boom and when everything starts to consolidate. All you gotta do is just find what % of profits is foreign in S&P 500, and look at what % of revenue is taken from overall GDP? You wouldn't have to bother with that correlation crap then.

 

I can see how buffett thinks it trails GDP growth though, since on average they grow at about the same pace. You really got to look by each 10 year period.

 

My guess is if you do this, you will find a fast growing % of profits from the S&P coming from overseas markets, and a growing revenue % number of US GDP after 1950.

 

Also it seems silly to compare results with dividends reinvested.

[constructive]

it seems to me ROE and ROIC will be better in emerging economies? There is less competition, demand outstrips supply more, less regulation and often the tax situation is better. And all those companies in their early stages have less bureacracy. Finally when looking at some of those emerging market companies, they often do business in other asian countries as well. And if there is only a small difference in ROE or ROIC, but your paying less then half, it takes a long time to catch up for the higher return company.

 

Historically emerging economies have had lower ROE than developed economies. That is why there has been a slight negative correlation between GDP growth and stock market returns.

 

Also you suggest that they have a tax advantage but tax rate is included in ROE.

[constructive]

Also it seems silly to compare results with dividends reinvested.

 

I think it would be silly not to. Would you calculate bond returns without including the yield?