Weight Loss, No Ozempic Required: Whither the US Share of Global Stocks?

By Victor Haghani and James White
ESTIMATED READING TIME: 5 min.

Just How Big Is the US Stock Market?

Ask MSCI or FTSE how much of the world stock market is American, and they’ll tell you about 62%. Ask the Securities Industry and Financial Markets Association (SIFMA), which reports raw, unadjusted market capitalization in its annual Capital Markets Fact Book, and the answer is 42%.¹

That 20-point gap comes from two adjustments the index providers make.

Free float. Index providers count only shares they judge to be available for trading. They exclude stakes held by governments, founding families, parent companies, cross-shareholders and insiders. Saudi Aramco for example is an extreme case, with the Saudi state owning most of it, so only a sliver enters the indexes.

Investability. Index providers also scale down or exclude companies and entire markets that foreign investors can’t easily access. The reasons include foreign ownership limits, capital controls, repatriation restrictions, and settlement or custody problems. China’s domestic A-shares, for example, enter MSCI’s indexes at a fraction of their size. Some markets are left out entirely.

These adjustments make sense for an index fund, since it can only buy what’s for sale, but they fall far more heavily on non-US markets. They tell you what you can buy today, not how the world’s listed equity capital is distributed.

Honey, I Shrunk the Index Weights

Over time, we’d expect the index providers’ 62% weight on US stocks to fall, because:

1. The US unadjusted market cap weight is 42% today.
2. The US accounts for roughly 26% of world GDP, and about 4% of world population. US public companies earn about 30% of global public-company earnings.² Market-cap weights tend to drift toward economic fundamentals, though erratically and on no fixed timetable.
3. Float and investability adjustments tend to shrink over time. Governments sell down stakes, cross-shareholdings unwind, ownership limits loosen, and markets get upgraded.
4. A weaker dollar would decrease the US weight, even if local prices don’t change. The IMF and major banks agree that the US dollar sits roughly 10% to 15% above its long-term Purchasing Power Parity.
5. Heavier US buybacks shrink the US weight; heavier US IPO activity would push the other way.

Bogle Was Right, Back Then

Vanguard founder John Bogle argued that US investors need only the S&P 500, because US companies earn so much of their profit abroad.

In the 1990s, when Bogle made that suggestion, giving up some diversification for simplicity may have been a sensible trade, but it isn’t anymore.

At the time, US and non-US stocks traded at similar price-earnings multiples, and thus had similar long-term expected returns. Today, US stocks trade at about 27 times trailing earnings and non-US stocks at about 16.³ A US-only investor now overpays twice:

• For the foreign earnings of US companies, relative to local companies’ earnings in the same markets.
• For the US earnings of US companies, relative to the US earnings of non-US companies.

Is My Home (Bias) Bigger than Yours?

US investors on average hold 15% of their stock exposure outside the US. Measured against the unadjusted non-US weight of 58%, that’s a very large home bias.

You might think home bias washes out if everyone has it in equal measure. But it doesn’t, at least not if everyone is overweight their home market by the same number of percentage points.

Here is a simple example. Say US stocks are 40% of world market cap, and the rest of the world is six markets of 10% each. Each country’s investors have wealth equal to their market’s size, so Americans have $40 to invest and non-US investors have $60 in total. US investors hold 85% at home, 45 points overweight. Investors in each of the other markets are also 45 points overweight, holding 55% at home. Everyone spreads the rest of their money in proportion to market cap.

Equal home bias, unequal outcome.

Investors in aggregate want $46 of US stocks, but only $40 exist.

The accounting is simple. Americans put $6 abroad, so foreigners can collectively own only $6 of US stocks, but they want $12.⁴ For supply to balance demand, the price of US stocks must rise relative to non-US ones.

Connecting the Dots

The index providers put a 62% weight on US stocks rather than the unadjusted weight of 42%, in part so that index investors don’t overwhelm foreign markets. The irony is that in a home-biased world, it does the opposite of what’s needed: it adds weight to the one market that’s already oversubscribed.

Meanwhile, the raw 42% tells you something about where index weights are headed. So does the US share of global earnings, GDP and population, all of which are lower still. Over long periods, float and investability adjustments tend to shrink, currencies tend to mean revert, and market caps tend drift toward fundamentals. Of course, changes in market prices tend to dominate in the short term, but in the long term, a lot of arrows are pointing in the same direction: toward a lower US weight in global indexes.

Home bias helps explain how the gap between US weight and fundamentals opened. The US is the world’s largest market, so equal home bias everywhere creates excess demand for US stocks. Some of the US’s outperformance over the past 30 years likely reflects that pressure pushing US valuations above everyone else’s. The relative multiple can’t expand forever, though it can keep expanding for longer than humans will still be making asset allocation decisions! But just because it can doesn’t mean it will.

We’ve seen this movie before. In 1989, Japan was more than 40% of the MSCI World index, larger than the US at the time and vastly more than its earnings or GDP weight. Today, Japan is just over 5%. In the first quarter of the 20th century, UK stocks were about 25% of global market capitalization, against 3.5% today.

A US-heavy investor now holds one of the world’s most expensive markets at a weight well above both its raw market cap and its economic footprint. None of this says when the weights will shift, they move slowly and can overshoot for years. But a portfolio with 85% in US stocks is a large bet that what can’t go on forever will go on a while longer. That bet should be sized deliberately: not by default, nor solely by the index provider’s weights, nor by decades-old advice to trade diversification for simplicity.

More Elm Articles on Global Diversification and US Home Bias

• There’s No Place Like Home: The Case For and Against Extreme Home Bias in Equity Investing
• Do US Industry-Sector Weights Explain the Higher Valuation of US vs non-US Equities?
• Introducing the Elm Partners All-Equity Portfolio For Non-US Investors
• Introducing P-CAPE: Improving Our Favorite Returns Estimator
• All Quiet on the Stock Market Front?
• Un-Exceptional US Stock Market Earnings?
• Concentrating on Concentration

1. We’ve adjusted SIFMA’s year-end figures for market moves since.
   
2. Based on Vanguard-reported P/Es of 27.0x for their broad US equity ETF (VTI) and 16.4x for their broad non-US equity ETF (VXUS), applied to SIFMA’s unadjusted market caps.
   
3. The same picture holds when using Payout-and-Cyclically-Adjusted Price Earnings ratios.
   
4. For supply to equal demand, each non-US market’s investors would need to hold only 25% of their neutral foreign allocation, just as US investors do (15% vs. 60%). That means 22.5% abroad (as 22.5% vs 90% is also 25%) and 77.5% at home, a 67.5-point overweight. For the market to clear, investors in small markets need far more home bias, in absolute terms, than investors in the US. If we’d set up this example with the US at 60% instead of 40%, and then 4 countries at 10% each for the balance, the small countries’ home bias would need to be even bigger relative to the US to clear.