Avoid US concentration risk with a GDP-weighted portfolio

[markus75]

Actually, I am a big fan of a market-capitalized index, such as the FTSE All-World ETF recommended here, which reflects the opinions of all market participants and weights it according to the absolute wisdom of the market.

But given the enormous increase in the weight of the USA, which now accounts for 65% of the All World index, I wonder whether the weighting of countries in the index should not be limited in order to control risk.

I would like to put forward the arguments against marketcap weight from a German blog post for discussion here, which is that if a country has too much weight in the index, a black swan event, for example, could have a great negative impact on the investor.

Blog article: https://gerd-kommer.de/marktkapitalisierung-vs-bip/
It's in german but you can use google translate:
https://gerd--kommer-de.translate.goog/ ... r_pto=wapp

My aim is not to map the world exactly according to GDP, which would be far too complicated, but to have a simple cap that limits the weighting of a particular country.

I my opinion, it would not require a particularly complicated portfolio to implement this.
The simplest would be a 3-Funds Portfolio:

30% S&P 500 or MSCI North America
40% MSCI World exUSA
30% MSCI Emerging Markets

With this portfolio, no country could be above 40%.

Opinions on this?

[ScubaHogg]

The USSR had a GDP

Do you think investing in it would have been a good idea?

[jg12345]

Actually, I am a big fan of a market-capitalized index, such as the FTSE All-World ETF recommended here, which reflects the opinions of all market participants and weights it according to the absolute wisdom of the market.

But given the enormous increase in the weight of the USA, which now accounts for 65% of the All World index, I wonder whether the weighting of countries in the index should not be limited in order to control risk.

I would like to put forward the arguments against marketcap weight from a German blog post for discussion here, which is that if a country has too much weight in the index, a black swan event, for example, could have a great negative impact on the investor.

Blog article: https://gerd-kommer.de/marktkapitalisierung-vs-bip/
It's in german but you can use google translate:
https://gerd--kommer-de.translate.goog/ ... r_pto=wapp

My aim is not to map the world exactly according to GDP, which would be far too complicated, but to have a simple cap that limits the weighting of a particular country.

I my opinion, it would not require a particularly complicated portfolio to implement this.
The simplest would be a 3-Funds Portfolio:

30% S&P 500 or MSCI North America
40% MSCI World exUSA
30% MSCI Emerging Markets

With this portfolio, no country could be above 40%.

Opinions on this?

I think the tacit assumption here is that the market is overvaluing the USA.

It might be or it might not be, I don't know.

But I don't know _better_ than the market either, so I am happy with market based.

[dcabler]

Actually, I am a big fan of a market-capitalized index, such as the FTSE All-World ETF recommended here, which reflects the opinions of all market participants and weights it according to the absolute wisdom of the market.

But given the enormous increase in the weight of the USA, which now accounts for 65% of the All World index, I wonder whether the weighting of countries in the index should not be limited in order to control risk.

I would like to put forward the arguments against marketcap weight from a German blog post for discussion here, which is that if a country has too much weight in the index, a black swan event, for example, could have a great negative impact on the investor.

Blog article: https://gerd-kommer.de/marktkapitalisierung-vs-bip/
It's in german but you can use google translate:
https://gerd--kommer-de.translate.goog/ ... r_pto=wapp

My aim is not to map the world exactly according to GDP, which would be far too complicated, but to have a simple cap that limits the weighting of a particular country.

I my opinion, it would not require a particularly complicated portfolio to implement this.
The simplest would be a 3-Funds Portfolio:

30% S&P 500 or MSCI North America
40% MSCI World exUSA
30% MSCI Emerging Markets

With this portfolio, no country could be above 40%.

Opinions on this?

I understand the concept, but I still have to wonder about the interconnectedness of worldwide markets. Can we really say that a black swan event that affects the 30% US share won't have a ripple effect into the 40% MSCI World exUSA, for example?

Cheers.

[Svensk Anga]

This looks like a system to overweight China. It has large GDP but pretty small amount of equity available in its public markets relative to the size of the enterprises.

Just because a place has substantial GDP, does not mean it will generate decent return to investors. Even if it does generate decent return now, maybe it has poor future prospects relative to alternatives. Cap weighting sorts all this out.

[markus75]

Can we really say that a black swan event that affects the 30% US share won't have a ripple effect into the 40% MSCI World exUSA, for example?

We'll never know what the next black swan event looks like. It could be like this way or it could be that way.

The article is argueing:
"Anyone who wants to make their portfolio "ultra-stable" also thinks about black swan risks, which may only occur every 20 to 40 years and which cannot be calculated or forecast. One of these "tail risks" is that a country with a particularly large stock market, due to its own internal failure or external reasons, gets into a sudden or gradual malaise that does not affect the other 194 countries in the world in the same way. This risk could exist for a passive stock investor today in relation to the USA, because the country has a weighting of around 65% in a market cap-weighted portfolio. A rational method of mitigating this risk and other forms of concentration risk in the portfolio is GDP weighting or a combination of GDP and MC weighting"

I think the goal of this “ultra stable” strategy is not to maximize returns, but rather to preserve wealth, which applies to conservative investors or retirees.

[typical.investor]

This looks like a system to overweight China. It has large GDP but pretty small amount of equity available in its public markets relative to the size of the enterprises.

Additionally, anyone with accurate information on China’s GDP would likely be jailed for releasing it. Russia may be the same though it’s no longer investible.

Cap weighting at least is what market participants have decided that investible assets are worth.

By all means, invest more in ex-US indexes than their market weight if you feel the U.S. market is overvalued, but using GDP as a guide ignores the reality that not everything is investible and GDP is prone to manipulation.

[manuvns]

US GDP has double from 2008 to now but SPY has balooned 6 times in same time period , China GDP has maybe gone 100 times but the stock market is heading down .

[rkhusky]

There are a lot of weightings better than GDP, but none likely better than market weight.

[JMACatfish]

Actually, I am a big fan of a market-capitalized index, such as the FTSE All-World ETF recommended here, which reflects the opinions of all market participants and weights it according to the absolute wisdom of the market.

But given the enormous increase in the weight of the USA, which now accounts for 65% of the All World index, I wonder whether the weighting of countries in the index should not be limited in order to control risk.

I would like to put forward the arguments against marketcap weight from a German blog post for discussion here, which is that if a country has too much weight in the index, a black swan event, for example, could have a great negative impact on the investor.

Blog article: https://gerd-kommer.de/marktkapitalisierung-vs-bip/
It's in german but you can use google translate:
https://gerd--kommer-de.translate.goog/ ... r_pto=wapp

My aim is not to map the world exactly according to GDP, which would be far too complicated, but to have a simple cap that limits the weighting of a particular country.

I my opinion, it would not require a particularly complicated portfolio to implement this.
The simplest would be a 3-Funds Portfolio:

30% S&P 500 or MSCI North America
40% MSCI World exUSA
30% MSCI Emerging Markets

With this portfolio, no country could be above 40%.

Opinions on this?

Not quite sure what to think.

On one hand, I think wanting to limit exposure to the US relative to market cap weighting is reasonable. On the other hand, I'm wary of the fact this article mostly justifies their advocacy for the GDP weighed portfolio with backtests and case studies. Not much mention of theory or underlying explanations beyond a quick nod to expected returns being higher.

I'm also not sure to what extent even a measure like GDP weighing would be sufficient for a black swan event. The 4 case studies listed were the Russian and Chinese Communist Revolutions, the rise of the Third Reich, and the stagnant Japanese stock market. For the lattermost case heavy global diversification across countries would've worked out very well, but otherwise? I think the whole concept of "if it goes to 0% you have bigger problems than your 401k" is a bit reductionist, but given the magnitude of most of these examples I believe it's a fair point. If you're an American and are concerned of regime collapse, perhaps the measures you should take go beyond diversifying your portfolio.

If your worries are confined to the stock market, an asset allocation around 30US/70EXUS is a pretty significant bet against current valuations of the US, but certainly one you could make. I buy into the idea of an efficient market pretty wholly and thus would personally be rather uneasy deviating from market cap by so much.

[pseudoiterative]

30% S&P 500 or MSCI North America
40% MSCI World exUSA
30% MSCI Emerging Markets

Suppose you establish a portfolio like this today, using a low-fee passively managed market-weighted ETF for each of the 3 components. Over time the weights are likely to change due to differences in relative performance. For example, suppose that after a few years your portfolio has ended up at 40% S&P 500 35% World ex USA 25% EM

What's your plan ("policy") for rebalancing?

A baseline approach could be to rebalance annually back to your target weights -- see Rational rebalancing: An analytical approach to multiasset portfolio rebalancing decisions and insights (vanguard research, 2022)

[dcabler]

Can we really say that a black swan event that affects the 30% US share won't have a ripple effect into the 40% MSCI World exUSA, for example?

We'll never know what the next black swan event looks like. It could be like this way or it could be that way.

The article is argueing:
"Anyone who wants to make their portfolio "ultra-stable" also thinks about black swan risks, which may only occur every 20 to 40 years and which cannot be calculated or forecast. One of these "tail risks" is that a country with a particularly large stock market, due to its own internal failure or external reasons, gets into a sudden or gradual malaise that does not affect the other 194 countries in the world in the same way. This risk could exist for a passive stock investor today in relation to the USA, because the country has a weighting of around 65% in a market cap-weighted portfolio. A rational method of mitigating this risk and other forms of concentration risk in the portfolio is GDP weighting or a combination of GDP and MC weighting"

I think the goal of this “ultra stable” strategy is not to maximize returns, but rather to preserve wealth, which applies to conservative investors or retirees.

It was pretty clear that this wasn't about maximizing returns but about improving diversification and reducing overall risk.

Regarding what the article is saying, yes, this makes sense for malaise. The issue, however, is that for many people, building such a portfolio country by country becomes unwieldy.

So something like what was suggested in the OP can at least be easily constructed and we can also look at some historical data.
30% S&P 500 or MSCI North America
40% MSCI World exUSA
30% MSCI Emerging Markets

First, looking at the correlations between each asset, using the Simba Spreadsheet which has data going back to 1976 (limited by Emerging markets). I see correlations as follows:
SP500-Int'l developed: 0.67
SP500-EM: 0.47
Int'l developed-EM: 0.61

Anything with correlations less than one would certainly improve diversification, though not as much as choices with 0.0 correlation.

When I look at the raw return data, however, I see plenty of examples in the annual history where both SP500 and International developed EX US both had double digit losses.
73/74 bear market
2001-2002 downturn
2008 of course for GFC (we can call it a black swan if we like)
2022 as well.

A few more cases on top of that where both had losses where one or both were single digit losses.

And it looks like EM joined in the fun as well, but not as often as either US or International developed has. And we can get a glimpse of that in the lower correlations at least between US and EM.

At least with the big markets, I still wonder about the interconnectedness of large markets, as defined by the example portfolio, on the downside whether we're talking about historical black swans or just everyday bad markets.

This can be quantified, if one wants to, by using something like the diversification ratio - where one takes into account the sum of weighted, individual volatility of each contributor relative to the volatility of the entire portfolio, but also takes into account their correlations. vineviz posted about this a few years ago and also provided a spreadsheet with the math to do the calculations - last I checked it's still out there if you can find the post he made with the link.

Cheers.

[nisiprius]

If you believe that matching the composition of the worlds' stock markets creates "US concentration risk," then it is rational enough want to reduce it.

But there's no reason to think that there's anything particularly "right" or optimum about GDP weighting.

[alex_686]

How would one build such a portfolio?

Nestle is headquartered in Switzerland but I think the plurality (note, not majority) of its revenue is in the US. Toyota is headquartered in Japan, most of its employees are outside of Japan, most of its sales are other countries. Don't get me started on Alphabet. And where do we slot Spotify? Market cap indexes can't figure out if they are American or European.

The USA will be underrepresented because it is dominated by multinationals. Germany would be weirdly represented - they are overrepresented with family businesses.

[SteadyOne]

Some countries like the U.S. and Japan have multinational corporations that derive substantial revenues from international operations. While their profits are reflected in stock market value, a significant portion - 30 /40% of their earnings isn’t part of the domestic GDP. So, the US stock market capitalization being larger than GDP. Should this portion of US market capitalization be ignored if weighted by GDP?

[jg12345]

Can we really say that a black swan event that affects the 30% US share won't have a ripple effect into the 40% MSCI World exUSA, for example?

We'll never know what the next black swan event looks like. It could be like this way or it could be that way.

The article is argueing:
"Anyone who wants to make their portfolio "ultra-stable" also thinks about black swan risks, which may only occur every 20 to 40 years and which cannot be calculated or forecast. One of these "tail risks" is that a country with a particularly large stock market, due to its own internal failure or external reasons, gets into a sudden or gradual malaise that does not affect the other 194 countries in the world in the same way. This risk could exist for a passive stock investor today in relation to the USA, because the country has a weighting of around 65% in a market cap-weighted portfolio. A rational method of mitigating this risk and other forms of concentration risk in the portfolio is GDP weighting or a combination of GDP and MC weighting"

I think the goal of this “ultra stable” strategy is not to maximize returns, but rather to preserve wealth, which applies to conservative investors or retirees.

I think the point made by other posters that US multinationals do large parts of their business elsewhere (and viceversa) makes this point moot.

"One of these "tail risks" is that a country with a particularly large stock market, due to its own internal failure or external reasons, gets into a sudden or gradual malaise that does not affect the other 194 countries in the world in the same way. " => if the USA gets into sudden or gradual malaise, Procter and Gamble Coca Cola Alphabet Meta etc etc will keep selling in other countries.

For a specific measure of that, I think that a third of all SP500 revenues are outside of US

[typical.investor]

I think the point made by other posters that US multinationals do large parts of their business elsewhere (and viceversa) makes this point moot.

"One of these "tail risks" is that a country with a particularly large stock market, due to its own internal failure or external reasons, gets into a sudden or gradual malaise that does not affect the other 194 countries in the world in the same way. " => if the USA gets into sudden or gradual malaise, Procter and Gamble Coca Cola Alphabet Meta etc etc will keep selling in other countries.


For a specific measure of that, I think that a third of all SP500 revenues are outside of US

It doesn't guarantee the US will always have access to those markets. If the free market rolls back and China make goods for China, the US for the US, Europe for Europe etc. It's obvious the pendulum is swinging back that direction and it is only a matter of how far it will go.

But should the worst come to pass, foreign investment might go out the window too the way it did in Russia. That's pretty far down the rabbit hole though.

In any case, I am not sure it makes sense to pay high valuations for SP500 companies to obtain revenue overseas. You could purchase foreign companies for a much lower price/earnings for their revenue at home and in the lucrative US as well.

[Valuethinker]

How would one build such a portfolio?

Nestle is headquartered in Switzerland but I think the plurality (note, not majority) of its revenue is in the US. Toyota is headquartered in Japan, most of its employees are outside of Japan, most of its sales are other countries. Don't get me started on Alphabet. And where do we slot Spotify? Market cap indexes can't figure out if they are American or European.

The USA will be underrepresented because it is dominated by multinationals. Germany would be weirdly represented - they are overrepresented with family businesses.

Switzerland is an excellent example.

Small country, population 8 million? GDP per capita is high.

But companies:

- Roche
- Nestle
- UBS
(probably a few others I have forgotten)

The ratio of market capitalisation to GDP must be very high.

Denmark also comes to mind because you've got 4m people? But Novo Nordisk, one of the world's largest healthcare companies (diabetes care). Orsted (offshore wind). Vestas (Wind turbines). (I can't remember if Carlsberg is still Danish listed?).

Sweden of course.

UK is also high and that's because c 70% of earnings of FTSE 100 companies are offshore. HSBC, BP, Shell, British American Tobacco, Imperial Tobacco, Diageo (Guinness + Johny Walker)

[ScubaHogg]

Worth saying again, but in plain language, that just because there is GDP doesn’t mean there are earnings. And earnings is what you are buying when you invest

[galeno]

I breached the same topic a while back. Basically I was trying to justify tilting more to Asia.

The USA is 4% of the world population but dominates the world equity indices in the 60+% range. It doesn't SEEM right. But it is.

[jg12345]

I think the point made by other posters that US multinationals do large parts of their business elsewhere (and viceversa) makes this point moot.

"One of these "tail risks" is that a country with a particularly large stock market, due to its own internal failure or external reasons, gets into a sudden or gradual malaise that does not affect the other 194 countries in the world in the same way. " => if the USA gets into sudden or gradual malaise, Procter and Gamble Coca Cola Alphabet Meta etc etc will keep selling in other countries.


For a specific measure of that, I think that a third of all SP500 revenues are outside of US

It doesn't guarantee the US will always have access to those markets. If the free market rolls back and China make goods for China, the US for the US, Europe for Europe etc. It's obvious the pendulum is swinging back that direction and it is only a matter of how far it will go.

But should the worst come to pass, foreign investment might go out the window too the way it did in Russia. That's pretty far down the rabbit hole though.

In any case, I am not sure it makes sense to pay high valuations for SP500 companies to obtain revenue overseas. You could purchase foreign companies for a much lower price/earnings for their revenue at home and in the lucrative US as well.

for the "I am not sure it makes sense to pay high valuations for SP500 companies to obtain revenue overseas" no one says one should buy only SP500 and no one says we buy US stocks (as part of an MSCI ACWI ETF or equivalent) 'to obtain revenue overseas'.

The point is to buy the market because the market is pricing stocks with a better knowledge than any of us. JPM, Blackrock, State Street etc etc all are forecasting for next 10 years larger gains for EU stocks than US stocks due to high valuations. so the bad consequences of high valuations are (or at least might be) priced in already.

[Valuethinker]

Some countries like the U.S. and Japan have multinational corporations that derive substantial revenues from international operations. While their profits are reflected in stock market value, a significant portion - 30 /40% of their earnings isn’t part of the domestic GDP. So, the US stock market capitalization being larger than GDP. Should this portion of US market capitalization be ignored if weighted by GDP?

Also Japanese companies this is often (not always) exports.

US companies this is often (not always) the earnings of overseas subsidiaries (that may manufacture locally, or in China etc).

Japan and Germany have unusually export oriented economies for the G8. (Canada and Australia, also - natural resource exports. And in Canada's case a very 2-way trade in automotive and automotive parts with the USA). US and UK it is much more services oriented (think of royalties Disney makes, say; or Microsoft earning cloud fees overseas).

Cars I don't think the US exports much *except* to Canada and Mexico (where it also imports a lot - it's a transborder industry and effectively a single market). Oil, where the US is the largest producer, it's also the largest consumer. Some refined products it exports to Europe, but others it imports. US leads in petrochemicals, fertilizer etc (the Gulf States are its main rivals). Food the US is the world's largest exporter from memory (Brazil number 2?).