The value of international stocks, in pictures

[vineviz]

I suspect that the unusually strong recent performance of the U.S. stock market relative to international stocks might lead to investors misremembering the diversification benefits of international stocks. Symptoms may include the beliefs that international stocks always underperform and that international stocks aren't necessary because "U.S. companies are global".

I thought a stroll down memory lane might be helpful. In my scrapbook I found these images from the 1970s, 1980s, 1990s, and 2000s showing the MSCI(ex. US) index and Ibbotson's large cap US stock index growing during each of those decades.

Enjoy.

[SimpleGift]

Appreciate the good graphical presentation, thanks. Some will always quibble with the choice of start/end dates, but decades are about as defensible and unbiased a choice as any, in my view.

An example of the power and effectiveness of simple data visualizations to perhaps break down preconceived ideas.

[Northern Flicker]

Thanks for sharing them. The MSCI EAFE index goes back to 1969. The emerging markets index is since the late 1980’s and only resembles the current index since the mid-90’s. I assume before about 1994 the graphs show EAFE data for ex-US?

[nisiprius]

Appreciate the good graphical presentation, thanks. Some will always quibble with the choice of start/end dates, but decades are about as defensible and unbiased a choice as any, in my view.

An example of the power and effectiveness of simple data visualizations to perhaps break down preconceived ideas.

I quibble with the use of a linear scale on the vertical axis.

[nisiprius]

The presentation is described as illustrating 1) the "value" of international stocks, and 2) the "diversification benefits."

What the presentation actually shows is that international stocks have performed quite differently from US stocks. They have, and by quite a lot. They have leapfrogged each other. It's also true that memories are short and recency is a problem. People may forget how well international stocks did in 2000-2009, just as during 2000-2009 they might have forgotten how poorly they did in 1990-1999. It's important to remember all this, particularly the leapfrogging.

But the fact that international stocks have performed differently in different decades does not automatically show that they would have produced a "diversification benefit." Sometimes different is just different.

This chart shows the hypothetical growth of two portfolios, with and without international stocks, over the years 1969-2017. One is a US-only 60/40 portfolio of SBBI "Large Stocks" (= the S&P 500 over this time period), and 40% SBBI "intermediate-term government bonds," the closest proxy for Total Bond. In the second, we've gone to a 50/50 US/international split; that is to say, 30% S&P 500, 30% EAFE, 40% bonds. I've used a log scale for the plot (even though Excel sucks so badly at plotting log scales). I've also corrected for inflation, which doesn't really matter much but the data I had happened to be inflation-corrected and it seemed silly to un-correct it.

This chart in no way contradicts the earlier charts. In fact it shows exactly the same leapfrogging effect. It is the same data, in the context of a portfolio as a whole (and with a log axis). (I hope I don't have to explain that the red line's being below the blue line most of the time is just the luck of the starting point. What we see is the global portfolio beating US-only for two decades, then the US closing the gap, then global pulling ahead for a decade, then US closing the gap... again).



Is there some diversification benefit to international stocks? Oh, probably, but I find it hard to muster much conviction about it. Diversification is only one of several forces tugging in different directions and it is debatable what the overall balance is. Is there a strong, robust, important benefit? Hmmm. The following curious table suggests some fragility of the benefit. It is taken from a book published in 2010, and it is part of a section that seeks to show incremental improvement, starting from a simple 60/40 portfolio and progressively layering on factor-based diversification. The author states that "the most important diversification on the equity side is to add an international exposure." Having stated that, he displays a comparison between portfolio 1, US only, and portfolio 2, with international exposure added. The table speaks for itself.

[azanon]

Are we comparing all S&P 500 vs. S&P 500/EAFE mix, or all S&P 500 vs. S&P 500/xUS stock (which would of course include emerging markets)?

linear to log graphs, now a complete axing of 20% of international stocks - the most uncorrelated to US stock group i might add. It's hard to keep up with the changes.

>Aside thought, if I found nisipirus' argument convincing, I certainly would pick right now to dump my international and go all US! Now if one has been all US stock since 09, i guess no big deal there because one has banked years and years of gains with every rebalance.

[pokebowl]

I've learned (from reading the hundreds of circular debates on this forum) the decision to either add or not add international equity comes down to personal investor bias. Figured I'd get this in on page one for the lurkers, prior to this thread exploding into 20+ pages of epic back and forth debates, filled with pretty charts and philosophical investing quotes.

[vineviz]

Thanks for sharing them. The MSCI EAFE index goes back to 1969. The emerging markets index is since the late 1980’s and only resembles the current index since the mid-90’s. I assume before about 1994 the graphs show EAFE data for ex-US?

For all periods the index I used is the MSCI WORLD ex USA index, which differs from the EAFE index slightly in that WORLD ex USA includes Canada but EAFE does not.

[nisiprius]

Are we comparing all S&P 500 vs. S&P 500/EAFE mix, or all S&P 500 vs. S&P 500/xUS stock (which would of course include emerging markets)?

linear to log graphs, now a complete axing of 20% of international stocks - the most uncorrelated to US stock group i might add. It's hard to keep up with the changes.

>Aside thought, if I found nisipirus' argument convincing, I certainly would pick right now to dump my international and go all US! Now if one has been all US stock since 09, i guess no big deal there because one has banked years and years of gains with every rebalance.

To be clear: my comparison was between 1) 60% S&P 500, 40% intermediate-term Treasuries, and 2) 30% S&P 500, 30% EAFE, 40% intermediate-term Treasuries.

Also, to be clear: I agree fully with (one of) vineviz's points, that there have been big differences in the decade-to-decade relative performance of US and international, and thus there will be important differences in the decade-to-decade returns of portfolios with and without international exposure. That's as close to a fact as anything can be, and I don't think it depends on details like which index was used.

EAFE goes back to 1969. It doesn't include emerging markets, because the term emerging markets hadn't even been coined then. The MSCI Emerging Markets index only goes back to IIRC 1988. We could include emerging markets but then we'd be throwing about 40% of the available data time period--and furthermore, it was a period of international outperformance. Look at vineviz's first chart. So the question is which is worse, throwing out 40% of available history or throwing out "20%" of international stocks.

However, if we want to use a global index, we can use PortfolioVisualizer. I used "backtested portfolio asset class" and chose asset classes similar to the ones I used for my chart. In order to make the colors match up with the chart I posted above, and with vineviz's charts, I chose to make the US-only portfolio "portfolio 2." "Portfolio 1" is the internationalized portfolio. The international index is:

Global ex-US Stock Market
AQR Global ex US MKT Factor (AQR Data Sets) 1986-1996
Vanguard Total International Stock Index Fund (VGTSX) 1997+

VGTSX definitely includes emerging markets. I can't actually find the AQR data set referenced; I assume it includes emerging markets.

As you see, during this thirty-year time period, the benefits of international exposure failed to overcome the handicaps.

Portfolio 2 (red)
Asset Class Allocation
US Large Cap 60.00%
Intermediate Term Treasury 40.00%

Portfolio 1 (blue)
Asset Class Allocation
US Large Cap 30.00%
Global ex-US Stock Market 30.00%
Intermediate Term Treasury 40.00%

Source

[vineviz]

What the presentation actually shows is that international stocks have performed quite differently from US stocks. They have, and by quite a lot.

Performing " quite differently" is, in fact, the ONLY condition necessary for diversification. Or, to be more precise, having less-than-100%-correlation.

I think the combination of charts I presented illustrates this in an interesting and non-technical manner, but I won't be offended if you disagree.

Is there some diversification benefit to international stocks? Oh, probably, but I find it hard to muster much conviction about it.

Many investors may be either too young or too old to remember the period from 1971 to 1988 (nearly two full decades) during which a globally diversified equity portfolio complex completely dominated a US-only equity portfolio.



In addition to the mathematical benefits that diversification provides, holding a well-balanced portfolio has been proven (time and again) to help protect investors from their own worse impulses. The 1970s and 1980s were a painful time for the "I don't need international stocks" investor, especially at the end: international stocks outperformed U.S. stocks four years in a row from 1985 to 1988 which appears to have led investors to chase performance afterwards (in 1989 and 1990, over 40% of equity mutual fund flows went into foreign stock funds, a spike from an average of just 6% in the prior four years). By then the dance was already over.

[SimpleGift]

To my mind, the dissection of past returns in this case has only limited information value about possible futures. About all we can say is that international and domestic stocks behaved differently in the past. Future returns are unknown.

Instead of the "past returns" argument, I'd like to highlight another one: "buy the haystack." The chart below shows the growth in the number of listed companies worldwide since 1975. While listed companies in the U.S. have been shrinking (in blue), the number of listed companies outside the U.S. has exploded in recent decades, especially in East Asia (in red).

NOTE: Only about one-third of the 45,000 companies shown below are actually "investable" today in the broad global indexes (14,400 companies, according to MSCI) — and these are concentrated in the U.S. and other developed markets, where the percentage of free-float shares and market capitalizations are much higher today.

• 
  Data source: World Bank

But if one is investing for the future, say the next 30-50 years, it might pay to notice where it is in the world today that entrepreneurial dynamism is greatest. The haystack is only growing larger beyond the shores of the United States.

[SimpleGift]

^^ Just to add: Not only has the number of listed companies outside the U.S. exploded since the 1970s, but the number of "investable" companies in the MSCI global indexes has been growing to keep pace (table below):

• 
  Source: MSCI
  

Looking ahead at the next 30-50 years, it's not a mystery where the vast majority of new companies in the investable haystack are going to be coming from (and it's not the U.S. in my view).

[GAAP]

But if one is investing for the future, say the next 30-50 years, it might pay to notice where it is in the world today that entrepreneurial dynamism is greatest. The haystack is only growing larger beyond the shores of the United States.

I'm curious if the average size of company in the USA has grown faster than elsewhere -- have we been consolidating faster, or just not adding new publicly-traded companies as rapidly?

Regardless, the trend of more growth elsewhere is not new, and I see no reason for that trend to stop.

[Pajamas]

I wonder how those charts look in Euros, rubles, francs, yuan, yen, etc., as well as in Venezuelan bolívars, gold, and barrels of oil. Plus U.S. housing!

That's the funniest thing I've come across all day.

[Northern Flicker]

Thanks for sharing them. The MSCI EAFE index goes back to 1969. The emerging markets index is since the late 1980’s and only resembles the current index since the mid-90’s. I assume before about 1994 the graphs show EAFE data for ex-US?

For all periods the index I used is the MSCI WORLD ex USA index, which differs from the EAFE index slightly in that WORLD ex USA includes Canada but EAFE does not.

Presumably that index started incorporating EM when the MSCI EM index was created in the late 1980's?

[SimpleGift]

But if one is investing for the future, say the next 30-50 years, it might pay to notice where it is in the world today that entrepreneurial dynamism is greatest. The haystack is only growing larger beyond the shores of the United States.

I'm curious if the average size of company in the USA has grown faster than elsewhere -- have we been consolidating faster, or just not adding new publicly-traded companies as rapidly?

Both have been happening in the U.S., as best I'm aware. Not only has the number of publicly-traded companies been decreasing in recent decades, but the remaining companies have gotten much larger, wealthier and more concentrated since 1975 (table below).

• 
  Source: Kahle & Stulz

And much more profitable, I might add, which helps account for their huge market capitalizations relative to the rest of the world.

[rich126]

Obviously we only have historical data to look at but I think it is foolish to think the future will continue to resemble the past. You have a rapidly aging US population (even more so if immigration is reduced), technology that is providing advantages to many countries, an education system that hasn't kept paced with the needs of a technological society, and a debt that is growing hugely.

I'm in the group that thinks it is foolish to put all your money in the US solely based on numbers from the past. Divide up the money and re-balance as necessary.

[Northern Flicker]

Many investors may be either too young or too old to remember the period from 1971 to 1988 (nearly two full decades) during which a globally diversified equity portfolio complex completely dominated a US-only equity portfolio.

From a secular perspective, I think there are 5 "performance" periods for US vs non-US equities since 1969, no doubt not described with perfect accuracy:

1. 1970's inflationary era with significant underperformance of US equities. US small-cap value greatly overperformed rest of US market 1977-1983. Bear market of 1982-1984 would demarcate the end of the period.

2. 1980's industrial boom in Japan and (West) Germany outperforming US (which also did well). Start of equity and real estate crash in Japan would demarcate the end of the period.

3. 1990's lost decade in Japan, emerging markets currency crises, and tech boom in US and other countries. US overperforms. Dot-com crash ended the period.

4. Early oughts (2002-2008) very large flow of capital to emerging markets economies. US underperformed. Global financial crisis ended the period.

5. Post-global financial crisis and superior recovery of US economy and US equity overperformance so far.

[azanon]

To be clear: my comparison was between 1) 60% S&P 500, 40% intermediate-term Treasuries, and 2) 30% S&P 500, 30% EAFE, 40% intermediate-term Treasuries.

It appeared to me that you were trying to make a rebuttal of the OP's claim where he was comparing/contrasting S&P 500 vs. Global xUS stocks, and then you used a straw man argument, of sorts, to show how EAFE index provided no real diversification benefit when mixed with S&P 500. I was just thinking at the time maybe that's true, but what did that have to do with the OP's point?

Don't get me wrong, I did find your original graphs interesting and very valuable (as I usually do with just about everything that you post). But that's besides my point.

As you see, during this thirty-year time period, the benefits of international exposure failed to overcome the handicaps.

There's no getting around the fact that if one wants to sing the praises of US stock out-performance, there's never been a better time to do it. That 30 years captures the 90s, and 09 to now.

Just a few years ago or so (I forget which year), one could show 30 year out-performance of bonds vs. stocks, and put 30 years in italics there too.

[vineviz]

Presumably that index started incorporating EM when the MSCI EM index was created in the late 1980's?

The name is a little confusing but it is still just developed markets. You can think of it as being pretty much just “EAFE+Canada” (which is yet another actual MSCI index, oddly).

[pascalwager]

I don't know the value of int'l stocks, but I also don't know the value of not investing in half of the stocks in the world stock markets.

[Index Fan]

I've learned (from reading the hundreds of circular debates on this forum) the decision to either add or not add international equity comes down to personal investor bias.

Actually, if diversification is a virtue (and it is always upheld by Bogleheads as such), one can say that less diversification isn't as good as more. This is a quantifiable, not a judgemental, factor.

Alright, didn't mean to get so serious Let the debate begin again!

[pop77]

How about comparing these portfolios with yearly rebalancing?

[vineviz]

How about comparing these portfolios with yearly rebalancing?

The graphs I posted are just one index vs another: there is nothing to rebalance.

[CULater]

Now, if Nisi can just take the currencies out of the returns. Is this much ado about the ebb and flow of currency differences? I expect mostly so.

[JoMoney]

International stocks...
Higher volatility
More explicit risks, in terms of sovereignty issues, currency risk etc..
Higher costs and tax implications
For better or worse, it isn't going to track with U.S. benchmark
If you recognize all that, and feel compelled that you need some allocation to it for the sake of "diversification"... go for it.

[MindBogler]

My position, given the fact that returns with / without international equity are close enough, hedging one's bet against a sustained equity downturn in the US is worth doing. It may very well come out behind in 20 years but we can't know what the future holds. Japan is the cliché example that is typically held up to justify an international exposure -- and for good reason.

[vineviz]

Obviously we only have historical data to look at but I think it is foolish to think the future will continue to resemble the past. You have a rapidly aging US population (even more so if immigration is reduced), technology that is providing advantages to many countries, an education system that hasn't kept paced with the needs of a technological society, and a debt that is growing hugely.

I'm in the group that thinks it is foolish to put all your money in the US solely based on numbers from the past. Divide up the money and re-balance as necessary.

You might be interested in the paper, if you haven't seen it before, entitled "Global Stock Markets in the Twentieth Century" by Philippe Jorion and William N. Goetzmann.

https://faculty.fuqua.duke.edu/~charvey ... Global.pdf

The did an admirable job of constructing a global index of stock returns, including countries whose markets dissolved or suspended trading. It paints quite an interesting picture of just HOW exceptional the American investing experience has been.

Yet, even including the complete collapse of several global stock markets it still turns out that over their study period a global equity portfolio had higher returns and lower volatility than a U.S. equity portfolio.

Over 1921 to 1996, the compound capital return on U.S. equities was 6.95 percent . . . . the return on the global comprehensive index was 7.25 percent.

[nisiprius]

Over 1921 to 1996, the compound capital return on U.S. equities was 6.95 percent . . . . the return on the global comprehensive index was 7.25 percent.

How is this to be reconciled with Dimson & al's findings that for 1900-2015, total real return (CAGR) was:

6.4% for the U.S.
5.0% for the world
4.3% for the world ex-US?

Is this just yet another example of endpoint dependence? If so, what is a rational basis for deciding whether to include or exclude 1900-1921, since apparently including or excluding it makes the difference between US > world and world > US?

And what do you make of their statement that (my boldfacing)

because we have no data on global market capitalization going back that far, we assign weights based on Gross Domestic Product?

[siamond]

To my mind, the dissection of past returns in this case has only limited information value about possible futures. About all we can say is that international and domestic stocks behaved differently in the past. Future returns are unknown.

Instead of the "past returns" argument, I'd like to highlight another one: "buy the haystack." The chart below shows the growth in the number of listed companies worldwide since 1975. While listed companies in the U.S. have been shrinking (in blue), the number of listed companies outside the U.S. has exploded in recent decades, especially in East Asia (in red).
[...]
But if one is investing for the future, say the next 30-50 years, it might pay to notice where it is in the world today that entrepreneurial dynamism is greatest. The haystack is only growing larger beyond the shores of the United States.

Exactly. Forget backtesting in this case, it doesn't help. Therefore hedge your bets for the future. Why is that complicated to understand? Beats me.

[siamond]

Over 1921 to 1996, the compound capital return on U.S. equities was 6.95 percent . . . . the return on the global comprehensive index was 7.25 percent.

How is this to be reconciled with Dimson & al's findings that for 1900-2015, total real return (CAGR) was:

6.4% for the U.S.
5.0% for the world
4.3% for the world ex-US?

Yeah, that's weird. I did a quick check, this doesn't seem to be fully explained by a start/end date issue (even if the 1921-1996 time period does make international look a bit better).

All the numbers I've seen so far are clear, the US premium (against International) was quite significant in the 1st half of the 20th century. For very obvious reasons, WW-I and WW-II totally destroying most of the other developed countries. A fact we should never forget when comparing US and International/global historical returns.

[fennewaldaj]

To my mind, the dissection of past returns in this case has only limited information value about possible futures. About all we can say is that international and domestic stocks behaved differently in the past. Future returns are unknown.

Instead of the "past returns" argument, I'd like to highlight another one: "buy the haystack." The chart below shows the growth in the number of listed companies worldwide since 1975. While listed companies in the U.S. have been shrinking (in blue), the number of listed companies outside the U.S. has exploded in recent decades, especially in East Asia (in red).
[...]
But if one is investing for the future, say the next 30-50 years, it might pay to notice where it is in the world today that entrepreneurial dynamism is greatest. The haystack is only growing larger beyond the shores of the United States.

Exactly. Forget backtesting in this case, it doesn't help. Therefore hedge your bets for the future. Why is that complicated to understand? Beats me.

Yeah it seems like there is a divide between those of us with the attitude that we can't know the future and therefore should invest with a split kinda close to market weight and those that use look at the not huge difference in returns and say it doesn't matter. To me it seem like the burden of proof is should be on the side that tries to justify not having significant international holdings. There is a significant contingent here who take the opposite position that the burden of proof should be on international to justify itself. So of course we look at the same data and come to opposite conclusions.

[jeffh19]

So what do all of you recommend for AA for US/international? 80/20?

In my brokerage, I’m only 11.7% international. I auto invest a good chunk of money each month, which I have set to 75% VTSAX 25% VTIAX. I’m thinking of maybe doing 50/50 for a while as international is supposed to be a better value right now, and I keep hearing not great things predicted for the Us market the next several years. I can always adjust back if the US market crashes to buy a ton more of that if it becomes a much better value with a correction or crash.

I’m also 100% US in my Roth IRA, and 13% international in my 401k.

Not trying to hijack the thread, but I’m curious what all of you guys above would recommend.

[JoMoney]

So what do all of you recommend for AA for US/international? 80/20?

In my brokerage, I’m only 11.7% international. I auto invest a good chunk of money each month, which I have set to 75% VTSAX 25% VTIAX. I’m thinking of maybe doing 50/50 for a while as international is supposed to be a better value right now, and I keep hearing not great things predicted for the Us market the next several years. I can always adjust back if the US market crashes to buy a ton more of that if it becomes a much better value with a correction or crash.

I’m also 100% US in my Roth IRA, and 13% international in my 401k.

Not trying to hijack the thread, but I’m curious what all of you guys above would recommend.

I don't have an international allocation, and I wouldn't suggest someone else do something I'm not doing.
That said, I'm not absolutely opposed to it, or think it's going to somehow be disastrous (even though I do think there's more risk involved, and higher expenses). Some suggest holding at global market-cap weightings, I (and Mr.Bogle) don't think it's necessary for U.S. investors to hold international. If you split the difference it probably puts you at about 20-25% of equities.

[drk]

So what do all of you recommend for AA for US/international? 80/20?

This is a topic that has been endlessly discussed on the forum.

[oldzey]

So what do all of you recommend for AA for US/international? 80/20?

In my brokerage, I’m only 11.7% international. I auto invest a good chunk of money each month, which I have set to 75% VTSAX 25% VTIAX. I’m thinking of maybe doing 50/50 for a while as international is supposed to be a better value right now, and I keep hearing not great things predicted for the Us market the next several years. I can always adjust back if the US market crashes to buy a ton more of that if it becomes a much better value with a correction or crash.

I’m also 100% US in my Roth IRA, and 13% international in my 401k.

Not trying to hijack the thread, but I’m curious what all of you guys above would recommend.

I live in the U.S. and have 0% international equities (or bonds) in my portfolio, which works for me.

[jhawktx]

International stocks...
Higher volatility
More explicit risks, in terms of sovereignty issues, currency risk etc..
Higher costs and tax implications
For better or worse, it isn't going to track with U.S. benchmark
If you recognize all that, and feel compelled that you need some allocation to it for the sake of "diversification"... go for it.

+1. You are paying more for the diversity you seek.

[akblizzard]

Is there some diversification benefit to international stocks? Oh, probably, but I find it hard to muster much conviction about it.

Thank you for these and past posts on this topic. I know they take time and effort, and I appreciate that. Everyone needs to decide what’s best for them and make a plan. Your posts have helped me make a plan I’m satisfied with.

[Northern Flicker]

Obviously we only have historical data to look at but I think it is foolish to think the future will continue to resemble the past. You have a rapidly aging US population (even more so if immigration is reduced), technology that is providing advantages to many countries, an education system that hasn't kept paced with the needs of a technological society, and a debt that is growing hugely.

I'm in the group that thinks it is foolish to put all your money in the US solely based on numbers from the past. Divide up the money and re-balance as necessary.

While I don't disagree with your conclusion, the problems you mention are not limited to the US and it is easy to have home bias against these types of risks in one's home country due to their familiarity.

Half of emerging market equity and most non-US developed markets equity is in countries with populations that tilt older than the US.

About 22% of the equity in a total int'l index fund is in countries with a higher sovereign debt-to-gdp ratio than the US (Japan, Italy, Belgium, Singapore, Portugal, Greece, Cyprus, Egypt).

And private (corporate and personal) debt as a %GDP is lower in the US than in China or most developed market countries (lower than all DM countries other than Japan, Canada, New Zealand).

[vineviz]

How is this to be reconciled with Dimson & al's findings that for 1900-2015, total real return (CAGR) was:

6.4% for the U.S.
5.0% for the world
4.3% for the world ex-US?

That link takes me to a Credit Suisse marketing brochure. I can’t find a peer-reviewed paper by those authors on this topic.

[siamond]

How is this to be reconciled with Dimson & al's findings that for 1900-2015, total real return (CAGR) was:

6.4% for the U.S.
5.0% for the world
4.3% for the world ex-US?

That link takes me to a Credit Suisse marketing brochure. I can’t find a peer-reviewed paper by those authors on this topic.

This was a reference to the 'Triumph of the Optimists' work (a seminal book published in Jan 2002 about global returns), and the continuation work published every year by the authors in the 'Credit Suisse Global Investment Returns Yearbook'. Here is a direct link to the 2016 yearbook.

Unfortunately, the 2017 and 2018 free yearbooks only come as an abbreviated version of the whole thing, while up to 2016, we had free access to the whole thing. This is THE reference for historical international returns.

[siamond]

Somebody asked me for a per-decade illustration of US (Total-Market, aka TSM) vs. the rest of the world (developed, aka EAFED). I added Emerging Markets (EM) for good measure. Data comes from a mix of Dimson and al. and what we have in the Simba backtesting spreadsheet.



Now do a little mental exercise. Project yourself at the end of each decade, one decade at a time. With zero knowledge of the future beyond this point. What conclusion would you reach? Be fair, try hard to set aside any preconceived idea.

For me, the conclusion is pretty clear: I would have been fooled a couple of times, I now realize that I have no clue where this is going, so I'll hedge my bets!

[CDub]

Somebody asked me for a per-decade illustration of US (Total-Market, aka TSM) vs. the rest of the world (developed, aka EAFED). I added Emerging Markets (EM) for good measure. Data comes from a mix of Dimson and al. and what we have in the Simba backtesting spreadsheet.



Now do a little mental exercise. Project yourself at the end of each decade, one decade at a time. With zero knowledge of the future beyond this point. What conclusion would you reach? Be fair, try hard to set aside any preconceived idea.

For me, the conclusion is pretty clear: I would have been fooled a couple of times, I now realize that I have no clue where this is going, so I'll hedge my bets!

Too true. This is a great mental exercise. There’s really no telling.

[nisiprius]

...

...

I think the inclusion of emerging markets is a distraction. It's fantasy football, for two reasons. I know Dimson & al tried to gen up some kind of numbers, but:

• the term "emerging markets" was not even coined until 1981; the first emerging markets index wasn't published until 1988. The Vanguard Emerging Markets Stock Index Fund had inception in 1994; I don't know if it was the first.
• When the index was created in 1988, emerging markets constituted only about 1% of market cap.

I don't think it makes much sense to think about whether you would have invested in emerging markets in the 1930s. The concept didn't exist, no vehicle for investing in it existed, and it would have been such a microscopic fraction of global market cap that adding it to a developed markets index fund--had such a thing existed, which it didn't--would have made essentially zero difference in performance.

[rnitz]

To me, the question is diversification to what? To little random performance movements up or down? No I don't need that. To major movements of currency value? Yes, that could have some value - if the dollar goes into the sh*tter for a large number of years then yes international diversification has a lot of value.

The real issue to me is do you want to bet on one country - the US. I live in the US, I've worked in the US my whole life, most of my assets are denominated in US currency, virtually all of my spending will be in the US, I personally think it's the greatest country in the world (no disrespect intended). But last century's US outperformance was largely based on the fact that major world wars devastated everyone else. If we have a major world war (heaven forbid) would the US be similarly spared this time?

International diversification for me is not for the little squiggle differences, it's for the black swans that have hit many, perhaps most other countries. I hope it's a wasted diversification (but regardless, I see it as at very little cost).

[siamond]

I think the inclusion of emerging markets is a distraction. [...] it would have been such a microscopic fraction of global market cap that adding it to a developed markets index fund--had such a thing existed, which it didn't--would have made essentially zero difference in performance.

I disagree about the lack of significance, but I agree that an EM data series isn't germane to THIS thread, and possibly distracting. We'll argue about it another time!

Here is the same chart, without EM, and extended to fully cover the 20th century, hence including the effect of the two world wars. You can click (once or twice) to see a larger version. Note that returns are nominals, and expressed in USD.

[JBTX]

I agree if returns are "different" seems to me there is diversification benefit. Looking at the OP graphs it is hard to conclude there is no diversification benefit.

We keep looking over limited data time frames which include the last 30-40 years where the US has had an overall extraordinary run. This seems to imply that this time frame is a reasonable representation of what to expect going forward. I couldn't disagree more. Any comparison of anything to the US stock market for the last 40 years is going to make the benefits of diversification look insigificant.

One only has to look at Japan as an example of the benefit of diversifying away from your home country. Sure that is a near worst case scenario, but the purpose of diversification is to lessen the impacts of worse case scenarios. In my mind this is more important than comparing standard deviations.

[Valuethinker]

So what do all of you recommend for AA for US/international? 80/20?

In my brokerage, I’m only 11.7% international. I auto invest a good chunk of money each month, which I have set to 75% VTSAX 25% VTIAX. I’m thinking of maybe doing 50/50 for a while as international is supposed to be a better value right now, and I keep hearing not great things predicted for the Us market the next several years. I can always adjust back if the US market crashes to buy a ton more of that if it becomes a much better value with a correction or crash.

I’m also 100% US in my Roth IRA, and 13% international in my 401k.

Not trying to hijack the thread, but I’m curious what all of you guys above would recommend.

I am not US based. It's crazy for me to underweight USA. I have, and it has hurt.

The best I can say to a US based investors is that Vanguard found the best volatility (risk) return tradeoff for a US investor was 30 per cent in foreign stocks (30 per cent of your equities). That seems to be a reasonable target weighting, rebalance yearly, say. That's based on historic experience.

You won't be too far wrong then if USA shoots ahead but if it is a decade of foreign outperformance, you will have exposure to it.

If you can't do all that with one fund, you could do 25 per cent developed market, 5 per cent Emerging Market.

Someone will perhaps post the link to the Vanguard study. I can't do so from where I am right now.

[zonto]

Someone will perhaps post the link to the Vanguard study. I can't do so from where I am right now.

https://www.vanguard.com/pdf/ISGGEB.pdf

Conclusion: In light of quantitative analysis and qualitative considerations, we have demonstrated that domestic investors should consider allocating part of their portfolios to international securities, and that a 20% allocation may be a reasonable starting point. Although finance theory dictates that an upper asset allocation limit should be based on the global market capitalization for international equities (currently approximately 51%), we have demonstrated that international allocations exceeding 40% have not historically added significant additional diversification benefits, particularly accounting for costs. For many investors, an allocation between 20% and 40% should be considered reasonable, given the historical benefits of diversification. Allocations closer to 40% may be suitable for those investors seeking to be closer to a market proportional weighting or for those who are hoping to obtain potentially greater diversification benefits and are less concerned with the potential risks and higher costs. On the other hand, allocations closer to 20% may be viewed as offering a greater balance among the benefits of diversification, the risks of currency volatility and higher U.S. to non-U.S. stock correlations, investor preferences, and costs.

[Valuethinker]

Someone will perhaps post the link to the Vanguard study. I can't do so from where I am right now.

https://www.vanguard.com/pdf/ISGGEB.pdf

Conclusion: In light of quantitative analysis and qualitative considerations, we have demonstrated that domestic investors should consider allocating part of their portfolios to international securities, and that a 20% allocation may be a reasonable starting point. Although finance theory dictates that an upper asset allocation limit should be based on the global market capitalization for international equities (currently approximately 51%), we have demonstrated that international allocations exceeding 40% have not historically added significant additional diversification benefits, particularly accounting for costs. For many investors, an allocation between 20% and 40% should be considered reasonable, given the historical benefits of diversification. Allocations closer to 40% may be suitable for those investors seeking to be closer to a market proportional weighting or for those who are hoping to obtain potentially greater diversification benefits and are less concerned with the potential risks and higher costs. On the other hand, allocations closer to 20% may be viewed as offering a greater balance among the benefits of diversification, the risks of currency volatility and higher U.S. to non-U.S. stock correlations, investor preferences, and costs.

Thank you.

Poster's suggestion of 20 per cent is therefore reasonable for the more risk averse investor.

Are the Vanguard numbers per cent of total portfolio or just of equities?

30 per cent of 60 per cent equities would be 18 per cent of total portfolio. If the 40 per cent is in US bonds then overall currency risk would probably be bearable.