Adding emerging markets to MSCI World spread the investment across more countries, but did little to soften the worst loss. From January 2001 to October 2026, in euros, the illustrative mixes raised compounded growth by up to 0.6 points a year after fund charges. The lead came from the first decade, and the worst fall was still close to 50%. The monthly correlation between MSCI Emerging Markets and MSCI World was 0.75, looser than the S&P 500's 0.97 or the 0.83 to 0.98 of five MSCI World factor indexes.
Emerging markets fell in all 24 of MSCI World's worst months. Before counting a second fund as diversification, multiply its portfolio weight by its country, sector and top-stock weights. You can see which exposures it adds now, even though the historical results cannot tell you which region will lead next.
Key takeaways
- Test the growth trade-off over more than one period. After fund charges, 80% MSCI World and 20% emerging markets grew at 6.9% a year from 2001 to October 2026, against 6.5% for World alone; from 2011 it grew at 10.6% against 11.8%. The result is not a forecast or an allocation target.
- The worst fall stayed close to 50%. MSCI World's deepest fall on month-end values was 49.9% after charges; with 10% to 30% in emerging markets it was 48.8% to 50.3%, and emerging markets alone lost 56.4%.
- The link is looser, but not in the worst months. A correlation of 0.75 is well below the S&P 500's 0.97, yet in MSCI World's 24 worst months emerging markets fell every time, by 8.6% on average against 8.5%.
- Compare rebalancing with letting weights drift. The 80/20 mix reset monthly grew at 6.87% against 6.78% for buy-and-hold over the full sample, but trailed from 2011, even before trading costs.
- Calculate what the fund adds to your holdings. On 30 October 2026, TSMC was 15.87% of MSCI Emerging Markets and information technology 41.77%. A fund at 10% of a portfolio with those weights would add about 1.6 points of TSMC exposure and 4.2 points of technology exposure.
How closely did emerging markets move with MSCI World?
Less closely than other equity indexes we measured, with a correlation of 0.75 over 310 months. MSCI World explained 56% of the monthly variance of emerging markets, against 94% for the S&P 500 in our S&P 500 study and 68% to 96% for the five factor indexes in our factor ETF study.
| Monthly net returns in EUR, MSCI Emerging Markets against MSCI World | Value |
|---|
| Correlation, January 2001 to October 2026 | 0.75 |
| Correlation, 2001 to 2010 / January 2011 to October 2026 | 0.85 / 0.65 |
| Rolling 36-month correlation: highest (March 2004) / lowest (April 2023) / October 2026 | 0.92 / 0.49 / 0.61 |
| Beta (move for each 1% move in MSCI World) | 0.99 |
| Months MSCI World fell / months emerging markets also fell | 113 / 83 |
| MSCI World's 24 worst months / emerging markets also fell | 24 / 24 |
Source: Copio calculation on MSCI Net Total Return EUR indexes; monthly levels and returns (CSV), rolling 36-month correlation (CSV).
The link loosened after 2010, but a beta of 0.99 means emerging markets still tended to move about one for one with MSCI World, and more widely around that average. A lower correlation with the same beta signals more independent swings in both directions, not a cushion.
Did emerging markets fall less when MSCI World fell?
Not reliably: emerging markets lost less in six of the eight MSCI World falls of more than 10% that we studied, and lost more in two.
MSCI World / MSCI Emerging Markets, in EUR: 2001-03 −47.1% / −35.7%; May 2007 to Feb 2009 −49.0% / −45.8%; 2011 −14.6% / −16.6%; 2015 −12.2% / −21.5%; 2018 −12.0% / −6.0%; 2020 −19.8% / −19.1%; 2022 −13.5% / −10.4%; 2025 −12.5% / −6.3%; 12 Oct to 18 Nov 2026 −3.6% / −10.8%. Source: MSCI index data (MSCI World 990100 and MSCI Emerging Markets 891800, net, EUR); Copio calculation; episode returns (CSV).
We selected eight episodes where MSCI World fell more than 10% on month-end values. In each we use the same dates for both indexes, starting at MSCI World's high for that window. The CSV gives the dates. The 2026 comparison is shorter and uses daily closes.
The financial crisis shows why the start date matters. From MSCI World's May 2007 high, emerging markets lost 45.8%, less than World's 49.0%. Start instead in October 2007, after emerging markets had rallied further, and the comparison reverses: 56.3% against 47.6%. In October 2008 alone emerging markets lost 19.5%, against 10.2% for World.
The current sell-off runs the other way from 2018, 2022 and 2025. From MSCI World's record close on 12 October 2026 to 18 November, it lost 3.6% in euros and emerging markets 10.8%, on daily index values. Our November 2026 AI sell-off analysis covers what happened inside MSCI World.
Did adding emerging markets raise compounded growth after costs?
Over the full period, yes, by 0.2 to 0.6 points a year; but all of it came from 2001 to 2010, and since 2011 each 10% in emerging markets cost about 0.6 points a year. The mixes below are illustrations, not recommended weights.
Growth a year after charges, Jan 2001 to Oct 2026 / 2001 to 2010 / Jan 2011 to Oct 2026: MSCI World only 6.5% / −1.5% / 11.8%; 90% World, 10% EM 6.7% / −0.2% / 11.2%; 80/20 6.9% / 1.2% / 10.6%; 70/30 7.0% / 2.5% / 10.0%; MSCI EM only 7.8% / 11.6% / 5.5%. Source: MSCI index data; charges from the iShares Core MSCI World factsheet (0.20%) and the Xtrackers MSCI Emerging Markets factsheet (0.18%); Copio calculation; mix results, gross and net (CSV).
Growth here is compounded wealth, the rate at which a balance actually grew, after deducting each fund's ongoing charge every month. A start in January 2001 favours emerging markets: MSCI World was at a month-end peak in euros, and in the decade that followed emerging markets beat it in nine calendar years out of ten. From 2011 to 2025 they beat it in five years out of fifteen (calendar years (CSV)).
Ten-year windows show how much the answer depends on dates. Of the 191 ten-year periods ending between December 2010 and October 2026, the 80/20 mix grew faster than MSCI World alone in 32%, and in none ending after October 2016. Its lead was 2.63 points a year for the ten years to December 2010 and its shortfall 1.61 points for the ten years to October 2022.
| After charges, EUR, Jan 2001 to Oct 2026 | World only | 90/10 | 80/20 | 70/30 | EM only |
|---|
| Volatility a year | 13.8% | 13.9% | 14.0% | 14.2% | 18.3% |
| Worst fall on month-end values | −49.9% | −48.8% | −49.5% | −50.3% | −56.4% |
| Worst 12 months | −39.5% | −39.2% | −40.5% | −41.9% | −51.0% |
| Longest time below a previous high | 146 months | 72 months | 64 months | 64 months | 81 months |
| €10,000 at the end of December 2000 became | €50,590 | €53,145 | €55,618 | €57,986 | €69,833 |
Source: Copio calculation; mix results (CSV).
The worst fall barely moved, and the worst 12 months got deeper above a 10% share. The longest time underwater shrank, but for a dated reason: MSCI World alone stayed below its January 2001 level until March 2013, while the mixes had regained their 2001 level by 2007 thanks to the emerging markets boom. Every mix then took five to six years, from its 2007 high to February 2013, to regain that high.
Did rebalancing beat buy-and-hold?
Only over the full period. Starting at the same 80/20 weights, the monthly reset compounded at 6.87% a year from 2001 to October 2026, against 6.78% for buy-and-hold. Starting in January 2011, resetting compounded at 10.64% against 10.92% for buy-and-hold. Both comparisons deduct the same fund charges and leave out trading costs and tax.
Resetting weights trims the market that has risen more and adds to the one that has lagged. That keeps exposures closer to the original target. It can also slow growth if one market keeps outperforming, as the 2011 comparison shows.
The 0.75 correlation cannot tell you whether resetting pays. In this sample, resetting once a year instead of monthly changed each mix's full-period growth by 0.1 point or less. Before trading to restore a target, compare how far your weights have drifted with what it costs you to reset them.
What does an emerging markets ETF hold in 2026?
A different index from the one that powered the 2000s: on 30 October 2026, Taiwan, South Korea and China made up 68.5% of it, and information technology 41.77%.
Weight in MSCI Emerging Markets, 31 Jul 2012 / 30 Oct 2020 / 30 Oct 2026: China 17.79% / 43.24% / 20%; Taiwan 10.68% / 12.67% / 28.88%; South Korea 15.47% / 11.9% / 19.66%; TSMC 1.94% / 5.71% / 15.87%; ten largest companies 16.54% / 32.45% / 37.54%; information technology 13.58% / 18.42% / 41.77%. Source: MSCI Emerging Markets Index (USD) factsheets of 30 October 2026, 30 October 2020 (archived) and 31 July 2012 (archived), page 2; weights (CSV).
In 2026, Samsung Electronics (6.99%) and SK Hynix (5.20%) follow TSMC among the largest companies, and India has grown from 8.15% in 2020 to 11.73%. Brazil, 13.07% in 2012, was 3.9%. The returns measured above came from an index with a different mix of countries and sectors, so they say little about how today's index will behave.
The index keeps changing. MSCI's November 2026 index review, announced on 11 November and implemented at the close of 30 November 2026, names Zhongji Innolight (China), Life Insurance Corporation of India and Nan Ya Printed Circuit Board (Taiwan) as the three largest additions.
For a World investor, that concentration adds to exposures already held. Our Asia article looks at the chip exposure inside Asian markets, and our S&P 500 study shows how to calculate a mix's look-through US share from factsheet weights.
Where does this leave an MSCI World investor?
An emerging markets fund is a decision about regions and sectors, not crash protection. It lowered the US share and added markets with their own swings. Those swings did not make regular rebalancing a consistent source of extra growth. It did not keep the worst fall away from 50%, and it fell in every one of MSCI World's 24 worst months.
Judged on compounded growth, the 25-year record supports neither a firm yes nor a firm no. The full-period gain rests on one decade that started from a favourable point; the last fifteen years show the opposite. A household that holds emerging markets needs reasons it can stand by through long stretches of trailing returns: from 2011, the shortfall has lasted fifteen years so far.
If the aim is to soften equity crashes, the candidates are holdings with a different driver. Our falling-months study and uncorrelated assets study measure which ones held up, and when they did not. For a portfolio meant to outlast the US's long lead, our long-term portfolio guide sets out the trade-offs.
Checks you can run on your own portfolio
- Write down your emerging markets fund's job. Higher expected growth, a smaller US share and crash protection are three different jobs; the data above supports the second, while the first and third depend on dates.
- Calculate your look-through shares. Multiply each fund's weight by its factsheet weights for Taiwan, China, South Korea and information technology, then add them up.
- Check the index name. MSCI Emerging Markets, MSCI EM IMI and FTSE Emerging differ in coverage; MSCI counts South Korea as emerging and FTSE Russell as developed, so mixing providers can double count or miss it.
- Stress your mix with the financial crisis. As a first pass, apply the common-window losses of 49.0% and 45.8% to your starting World and emerging-market weights. Compare the resulting loss in euros with the cash you could need. This estimates a hold-through loss for that window, not your portfolio's worst drawdown.
- Look at 2011 to 2026 before deciding. Ask whether you would have kept the fund through fifteen years in which emerging markets compounded at 5.5% a year after charges, against 11.8% for MSCI World.
- Check what a reset would cost. Compare the weights you own today with the exposures you intend to hold. Resetting reduces drift, but it lagged buy-and-hold from 2011 before accounting for spreads and taxes.
Listed infrastructure had the same 0.75 monthly correlation with MSCI World from 2004 to October 2026 and fell in 23 of its 24 worst months, as our infrastructure ETF study shows.
FAQ
Does an emerging markets ETF diversify an MSCI World ETF?
Partly. From January 2001 to October 2026, in euros, the monthly returns of MSCI Emerging Markets and MSCI World had a correlation of 0.75, a looser link than the S&P 500's 0.97. But emerging markets fell in all 24 of MSCI World's worst months, and a mix with 10% to 30% in emerging markets had a worst fall of 48.8% to 50.3% after fund charges, against 49.9% for MSCI World alone (Copio calculation).
Did adding emerging markets improve long-term returns?
Over the whole period, slightly: after charges, 80% MSCI World and 20% MSCI Emerging Markets, rebalanced monthly, grew at 6.9% a year from January 2001 to October 2026, against 6.5% for MSCI World alone. All of that came from 2001 to 2010. From January 2011 to October 2026 the same mix grew at 10.6% a year against 11.8%, and it was ahead in none of the ten-year periods ending after October 2016.
Did emerging markets fall less than MSCI World in crashes?
Sometimes less, sometimes more. Across eight selected falls of more than 10% in MSCI World since 2001, emerging markets fell less in six and more in two. In the financial crisis, from May 2007 to February 2009, they lost 45.8% against 49.0% for World, in euros on month-end values. From MSCI World's record on 12 October 2026 to 18 November, emerging markets lost 10.8% against 3.6%.
What does an emerging markets ETF hold in 2026?
Asia dominates, and technology is the biggest sector. On 30 October 2026, Taiwan was 28.88% of the MSCI Emerging Markets Index, China 20%, South Korea 19.66% and India 11.73%, according to MSCI's factsheet. Taiwan Semiconductor Manufacturing (TSMC) alone was 15.87% and information technology 41.77%. In October 2020, China was 43.24% and technology 18.42%.
Is an MSCI ACWI fund the same as MSCI World plus emerging markets?
It holds emerging markets at their market weight, which was 11.2% of the MSCI ACWI on 30 October 2026, and changes that weight with market values instead of resetting it. From January 2001 to October 2026 its monthly returns had a correlation of 0.996 with MSCI World, and it grew at 6.4% a year after a 0.20% charge, against 6.5% for an MSCI World fund (Copio calculation).