Deciding you should invest outside of the UK but then electing to put all your money in China is a case of out of the frying pan and into the wok.
It doesn’t spread your risk, and it exposes you to the biggest fear that most of us have when we make an investment, which is the potential for an all-out loss.
Studies have shown that as a species we prefer two birds in the hand to a potential five in the bush – if they come at the risk of a dead parrot.
In other words, we’re more averse to loss than we’re greedy for gains.
And that’s important in the context of overseas investing, because some countries have done far better or (more scarily) far worse over the long-term.
Long term returns from different countries’ stock markets
|Annualized real return||Cumulative since 19001||Annualized real return||Cumulative since 19002|
Small differences in returns matter
This cumulative real return data for each country was a real eye-opener for me the first time I saw it.
It’s a reminder that seemingly small differences have a major impact when it comes to compound interest.
In terms of annual return, the difference between investing in shares in the U.K. versus the U.S. doesn’t look like much:
- Averaged over the past 114 years, the annualized real return from equities5 for a British investor is 5.3%.
- Over the same period, U.S. investors enjoyed a slightly higher annualised return of 6.5%.
What’s 1.2% between two countries divided by a common language, you say?
Well, over the long-term such small differences really do add up.
- A U.K. investor who reinvested all her dividends since 1900 would have multiplied her portfolio 372 times over.
- A similar US portfolio would have multiplied 1,248 times!
And these are two countries where returns have been in the same ballpark.
In contrast a Swiss investor in its lower-returning equities would have seen his money multiply a mere 137 times, while an Italian Rip Van Winkle (and a bit) would have only seen his funds increase nine-fold after more than a century.6
Few people live to 114 of course, and none so far have been wizened old investors. Some will say that looking at returns over such a very long period is therefore misleading.
I disagree – provided you’re not using the data for more than what’s reasonable.
As a way of seeing how different countries have produced very different long-term returns, it’s perfectly useful.
But the data shouldn’t be used as a basis for cherry-picking one country over another when deciding how to allocate your money for the future.
Rather, it reinforces the case for diversifying very widely across global equities, which most people should do using index trackers.
Not one world stock market (yet)
Anyway, studying the data, it’s clear the countries that did best over the past century were those on the winning side of World War 2 – provided of course they did not “win” as communists, like the Russians and the Chinese.
Even before then, the value of Germany’s bonds was utterly destroyed by its rampant inflation prior to the war years.
Also realise that countries like Australia, New Zealand, and even the U.S.A. were still emerging markets in 1900. That’s another factor explaining their incredible performance (aside from their being out of the range of jackboots and bombers).
Some would argue that stock markets are now too closely correlated for this historic data to be of much interest.
I say – not so fast!
Firstly, I think these historic returns are still a highly-relevant pointer to the likely divergent outcomes from investing in different emerging markets today.
Secondly, we are still seeing very varied returns in just the past few decades. Look at the dire returns from Japan since 1989, for example, which makes even our miserable decade from 2000 to 2010 look like a hiccup.
You only have to consider the turmoil in Europe or more recently emerging markets to see that history hasn’t ended, and that a Star Trek-style United Nations of globalized ISA investors is probably still some way off.
In Fooled by Randomness, author and Black Swan-spotter Nassim Taleb points out that an investor in Russian or Chinese companies at the start of the 20th Century who suffered a complete wipe-out would have a rather different tale to tell about ‘investing for the long term’ to the Americans who typically write all the investing books.
Who is to say that the 21st Century won’t hold similar surprises?
Accordingly, I think spreading your money across world stock markets remains a good idea to reduce the risk of being 100% in an all-out lemon for 40 years, as well as for the more general diversification benefits.
Note: This article was updated in February 2014. Comments below may refer to previous data, but in most cases they are still relevant and interesting. Especially mine. 😉
- E.g. $1 invested in Australian equities in 1900 would be worth $3,332 in today’s money, if you’d reinvested all income. [↩]
- E.g. $1 invested in Australian bonds in 1900 would be worth $5.7 in today’s money, if you’d reinvested all income. [↩]
- All China figures are since 1993 only. [↩]
- All Russia figures are since 1995 only. [↩]
- The real return is the return after inflation has been taken into account. [↩]
- Note that all these numbers are all measured in US dollars and so will vary from year-to-year with the exchange rate. By using US dollars as a consistent currency, the Yearbook authors enable comparison across all the countries they cover. The prevailing exchange rate is therefore another factor to consider when comparing one country’s gains with another’s. [↩]
Couldn’t agree more, diversification is the key. What’s the best way to diversify across countries? Should I buy individual shares/ bonds/ index trackers/ invest in multinational firms?
Wow, this is huge. The figure for Australia is massive. A simple tracker fund there invested at any time during the last century would be worth a massive fortune now. I think of Australia as an advanced industrial economy not the third world, so how are the returns so big? And is it too late to buy into that success? Should we be putting money into the next big thing? Or maybe a tracker fund that includes lots of the Pacific nations?
Also, is there a way to jump on board the Sth Africa or emerging African nations train? Any tracker funds that follow their success?
@teamdave — As I mention in the piece, Australia was effectively an emerging market 111 years ago, in my opinion. Countries that have now faded like Argentina were arguably far more developed back then. Perhaps the lesson is (again!) to spread your money about — and to hold on for the long term. (I assume you’ve got a century to wait. I’m certainly popping the Vitamin C pills… 😉 ).
@MMR — For most people index funds can give exposure to emerging markets at a low cost. Or I like the Templeton Emerging Market Investment Trust (which I hold) if you believe that emerging markets are still inefficient enough to allow managers a reasonable chance to outperform.
1. “we’re more averse to loss than we’re greedy for gains”
careful here, are you a person looking forward or looking backwards in time when you say that?
Psychologists tend to agree and studies support that the pain of losing £10 is more than the pleasure of making £10 – but thats how we feel ‘after the fact’. Thats not the same as saying that if we had £10 to invest we are more likely to go for a low-risk low-return option than a high-risk high-return option. I *don’t think* that studies support the fact that we are smart enough to link these two viewpoints together in a rational way, especially if we have little real experience of loss.
2. quoting compound returns over time-spans far greater than anyone can hope to invest for is poor practice and misleading
other than that – great stuff 😉
I’d be interested to see the correlation there between natural resource abundance and return… might get back to you on that one.
@Ben, thanks for you feedback.
1) I am saying that it will hurt a lot if you put all your money in Norwegian (or Japanese) stocks and the market sinks for 30 years; that in that circumstance you are likely to be very unhappy, and unconsoled by “what you could have won”, Bullseye-style. It’s true it’s a little theoretical, since you’ll never go down the trouser leg of time where the Norwegian market quintupled in a decade in order to compare notes. 😉
2) Disagree. The returns over the period are what they are. Like the LBS professors, I think they put the tremendous divergence between different markets into perspective. Quoting any arbitrary 10/20/30 year period would be equally (more? less?) misleading in my view, especially as it would invariably be the past 10/20/30 years.
The takeaway point isn’t “Australia has gone up a lot so get exposure to that”. The takeaway is “markets do a lot different stuff, so unless you are very confident of your abilities / don’t care / are prepared to be wrong, then get exposure to (close to) them all”.
@Rob — True, but again there’s an emerging market element to that. If you could have invested in Britain just before we cut down all our trees to build the British navy, I’m sure that correlated to a great deal of wealth, too. Natural resources are one of the great conundrums of our time (15 years ago commodities had been all but written off as an investment theme after declining real prices for many, many decades). Personally I’d prefer to invest higher up the food chain if I was going to take an active approach (i.e. Invest in Apple over apples! (At the right price…) ).
1. TI, as ever, thanks for this. Very useful.
2. Those annualized real returns would be even more helpful if we had some idea of the ‘spread’ of the distribution of returns for each country: standard deviation, for example. That way, we could quantify the variability around the mean returns – providing a handle on the consistency of the returns year-by-year. This, of course, is your point about Japan (and others)…
TI, does the real return take into account currency differentials? Example. S Africa has returned 7.2% against the USAs 6.2%. However, 50 yrs ago the Rand and $ were 1:1, but now it takes say R8 to purchase 1$.
@Roger — They are in local currency terms. They are ‘real’ returns, too, deflated as far as I can see by local inflation.
@Alex — There’s no Standard Deviation data in the Yearbook; there might be in the bigger (and costlier!) Sourcebook. As mentioned above, though, I would be cautious about data mining this sort of data too hard. I don’t think anyone is going to discover that country X has the greatest returns for the lowest volatility *and therefore* the same will hold true in the future. I don’t think it would tell you that.
If I’ve misunderstood your question and you’re just saying it’d be interesting to see for evidence again that you typically pay for higher returns by enduring more volatility, I agree it’d be interesting to see.
1. TI, no, I don’t want to do the “data mining” you describe. I understand that past performance is not a reliable indicator of future performance. 🙂
2. My point only was that the mean doesn’t tell us anything about the spread of the distribution of returns since 1900 for each country. Your table allows cross-country comparison of mean returns only: the standard deviations would be complementary information.
One of the problems here is how do you actually make this diversification happen? and how do you weight it? for instance, in my pension AVCs I have a L&G Global Equity/UK 50:50 fund which looks pretty much like a developed world fund to me, and in my ISA I have a holding of LGAAAK which is a emerging markets tracker to balance it with some non-developed world stuff. However, looking at the AVC fund the top ten components look like a UK HYP, indeed I hold some as shares in my ISA. Which isn’t quite what I expected from the name. I only have the choice of two funds in the AVC and that was the most diversified.
A world diversified index fund would presumably be heavy on US stocks if weighted by market cap. So the mechanics of how to do this global diversification and actually *get* diversity aren’t clear to me, it doesn’t seem as simple as buy global index fund regularly, sit back, and wait 😉
@The Investor – agree with you fully there, but if we are talking about choosing a market to invest in… the US, Australia and South Africa jumped out at me as natural resources rich. I’ve done some digging (just in the last 5 minutes) and there isn’t really a clear correlation for oil resources, see the oil map of the world:
But move to what I suspected was a pattern with natural minerals (which are very abundant in South Africa and Australia) and see this:
Now going onto annualised real returns, the pattern is far less distinct, but in absolute terms its pretty clear, especially in placing countries higher than you might expect. A good economy would appear to be the dominant factor, but for their simplicity to understand if nothing else, natural resources would appear to be a good signal for investing in a countries index. I can’t help but find S.Africa’s annualised returns of 1.8% impressive against the UK’s 1.5%.
In real returns, traditionally strong economies do better, take Sweden for example with 2.2% real returns, but their greatest enemy would appear to be losing a world war: look at the real returns of Germany, Italy and Japan (though maybe Japan’s a bit special). Fascinating stuff… great article.
Interesting post, thanks. I see advisors recommending investing in specific countries, which the spate of country-specific ETFs now facilitate. But I think your return statistics demonstrate the risk. Why would country-picking be any easier or more profitable than stock-picking? I think it can be argued that countries are even less transparent and predictable than corporations! Agreed, we can plan on more surprises….
@ Ermine – a world diversified fund should be about 50% US as that’s how the world has diversified across global markets. If you wanted to second-guess that and dilute that weighting then you could buy specific country ETFs or an emerging market fund or whoever else you wanted to back.
@ TI – I think Ben has a point about realistic investing lifespans and it would be interesting to see how the annualised real return of the UK and USA, for example, played out over 20 or 30 years.
@ermine–One way to get world diversification without buying too heavily into they US would be to buy a portfolio of regional index funds and an emerging market index fund and weight them roughly evenly. That will increase the world wide spread of your investments but comes at the expense of having more funds to juggle.
just about half way through Shiller ‘Irrational Exuberance’ theres some stuff in a similar vein, particularly best/worst 1/5years for a whole slew of indices and what happened over the following adjacent time-span. Some of it is absolutely eye-watering, e.g. Philipines Nov1984-Nov1989 +1253.2% or even UK Nov1973-Nov1974 -63.3%. Theres some serious risk out there!
I don’t understand the figures.
How could say Italy with an annualised return of 1.7% only return a total of 6% after 111 years?
@TA Isn’t the point of this article that it would be advantageous to average weighted by discrete market (a little bit as Smorgasbord indicates?) as different areas do well at different times? It’s fair enough to weight your stock market investment by market cap, ie 50% US and then the rest as one viewpoint.
However, the indication seems that most of the growth comes from emerging markets, but you can’t tell which one, and may get slaughtered in some. So an equal weighting my market will be heavier weighted to EMs compared to by capitalisation. Weighting by capitalisation is but one viewpoint, it’s the obvious one in some ways but it will inevitably skew your holdings towards developed world elephants. I plan to use index funds for the non-UK part of my holdings but I’m not yet sure that weighting by capitalisation is the only, or even the rational, choice.
@ermine – prob need to factor in currency risk as well which will skew you towards a home bias – my feeling is its an inexact science, something along the lines of Hales suggestion would seem to be adequate
PS the UK recovered +73.0% the following year (Nov1974-Nov1975) – Nuts!
Chaps, you need to read the figures more closely. 🙂
@David — The 1.7% is a percentage return, but as explained in the copy the ‘6’ is a six-fold increase, in real terms. That is to say, if you’d invested 1000 Lira in the Italian market 111 years ago, you’d have 6000 Lira now.
@Rob — You’re looking at the bond column. The real equity return for the UK on an annualized basis is 5.2%. For South Africa it’s 7.2%.
Sorry can’t comment more today, up to my ears in work. Glad everyone finds the data interesting, but agree as with everything it raises many questions / conundrums even as it sheds light. 🙂
I maintain that the takeaway for most readers here is that they should bolt on some global equity exposure. I suspect more precision after that will only be proven right or wrong long after the fact (many decades after…)
@TA — It’s true that nobody will live for 111-years, but if anything I think this research shows even equity bulls how dangerous it is to just plop 5% into a calculator (even if we have no realistic alternative) for a particular market. Anyone who wants to do so is heartily encouraged to have a play with the Monevator compound interest calculator. 🙂
It’s a shame they don’t provide standard deviation / volatility information here, to this point and to Alex’s. If I can find that data somewhere else, I’ll get it up.
I’d urge everyone to remember that there have been times when equities have provided a negative real return over multiple decades. We even saw a negative number in the UK to 2010:
That’s a bit of an embarrassing mistake on my part (hangs head in shame). Enhances my point though I should point out.
@TI – very good point – Shiller is telling me the same – stocks cannot be considered risk-less simply because you have a 30 year investment time frame; ‘best for the long run’ needs to be treated with caution.
PS i think i mentioned before that geneticists believe the 1st 1000 year old has already been born – so maybe your lengthy compound returns would be of interest to him or her (or it)?
Ben, which “geneticists” have said that and where?!!
@alex – Cambridge University geneticist Aubrey de Grey (http://www.dailygalaxy.com/my_weblog/2009/03/can-humans-live.html). To be fair he looks like he may be verging on 1000 years old already… (http://en.wikipedia.org/wiki/Aubrey_de_Grey)
1. Thanks, Ben. His is an ‘unconventional’ position. 🙂
2. For accuracy, he’s not based at the University of Cambridge: he left employment there in 2006 (source: his CV at http://www.sens.org – and the Wikipedia entry you cite).
3. The first link you provide is dated March 27 2009 – and begins: “Cambridge University geneticist Aubrey de Grey has famously stated…” Hmm…
4. Well, it generates good headlines… 🙂
Warning over “negative returns over multiple decades” is really going too far. Negative decades are very rare and certainly not random. Putting aside the effects of two World Wars (I think we can agree they were exceptional) the US market only shows two episodes of losses in excess of 10 years, and these were both at the end of periods of fairly wild stockmarket excess. The late 60s and the late 90s. Even investments made in 1929 were back up after 7 years. The US market has never seen a two decade loss.
Note: these real total returns based on annual average figures, not spot highs and lows. Data from Shiller.
@Paul: you wouldn’t be saying that if you were Japanese. The s&p dropped 80.6% from sep 1929 to June 1932. The real s&p index (adjusted for inflation) didn’t return to it’s sep 1929 value until 1958. A period of 29 years. Source – Page 9, irrational exuberance, shiller
@Paul — I think the warning is valid. In the decade just gone to 2010 US real returns were negative. (http://monevator.com/2010/03/10/us-historical-asset-class-returns/). Japan is two decades out for the count now. There are others. I am about as bullish a blogger on equities you’ll find, and have been throughout this past three years when others have been screaming for the end of civilization. (E.g. http://monevator.com/2009/06/10/ten-year-bear-market/)
Your comment about negative periods following “fairly wild stockmarket excess” is of course correct. I for one would eat my hat and my shoes if the next decade saw negative real returns.
The US is a rather unusually successful stock market, for whatever reason. Perhaps as a result, its investment mantras tend to echo around the world. But many markets have not been anything like as successful.
Also, do we say that American exceptionalism means we shouldn’t *ever* fear/allow for an unusually prolonged period of negative returns, based on its fine track record? Or do we suspect that its fine run coincided with the US economic and military might scaling heights never seen before in any one country in the world — a trick that it’s easy to argue won’t be repeated in the next century, even if it muddles along fairly successfully like Britain when she lost her Empire?
Personally, I’d hedge my bets on all the above.
@paul -apologies didn’t read your post carefully enough spot values bs average values. So I was comparing apples against your oranges
The Japanese situation really makes my point for me…it was, and is, the result of a truly massive bubble, probably the biggest ever, justified at the time by “Japan’s different”. Seriously bad stock market performances only follow uncontrolled booms, they are not random events.
The current US etc “lost decade +” is the follow-on to the 9os excesses and well deserved.
The US long-term performance of 6.2% real goes back to 1870 (Shiller) and according to Jeremy Siegel (Stocks for the Long Run) to 1800, long before the US hegemony. I suspect that the difference between the US and other markets is that the US larely escaped the devastation of the two World Wars as well as bouts of socialism.
My analysis is based on the US data that Robert Shiller generously publishes on his website. I don’t know of any similar data for the UK market unfortunately or I would try that too.
@Ben, Using annual data does smooth out the pain somewhat. The Sept 29 Index was 20% higher than the 1929 average!
Please be assured I am not an unreconstructed bull. I sat out the market from 1998 to 2008 when the Shiller data convinced me that we are sufficiently under the trend line to make stocks a good long term prospect again. I know we have to say past performance is no guide………etc but personally I am happy to take past performance that has a +140 year track record.
@Paul — Thanks for the extra information and thoughts. I think we agree far, far more than we disagree; we’re really just arguing the toss over the ‘s’ in decades!
@ Ermine – the point of the article is to diversify internationally. The methodology is left to the beholder. While emerging markets are a source of much economic growth I haven’t seen any evidence that they’re the source of most stock market returns. Obviously they’re an asset class that’s performed well in recent years – and got battered in 2011. If you skew to EM then you’re certainly piling on the risk and paying a pretty high price to get in on the act. Those elephants may surprise you yet. Check out the UK number. No country has more declined in relative importance over the period yet those stock returns have held up pretty well.
The best benefit of course is to diversify your portfolio over a number of international markets. Now, this isn’t easily possible directly, but the best way to do it could be through a global ETF or something. I believe it was Solnik (1974?) who found that this resulted in the same average rate of return but at only HALF the risk due to the diversification involved. The really interesting thing is that the marginal benefit of this diversification became 0 after about 20 investments.
Interesting stuff. Of course very tempting to ‘backtest’ assumptions about why countries did well/badly. Factos mentioned above include emerging market & lots of remaining natural resources. In one sense US & Oz in 1900 were EMs, but the US was already the largest economy and Oz was already one of the richest in the world per capita. Clearly they still continued to grow, though…
Other factors I wonder about:
– Population growth (domestic & immigration) – all of the top countries had high immigration over the period.
– Economic freedom – top countries had high market freedom, interestingly followed by Scandinavian countries which have low inequality and more socialistic.
– Worse performers often affected by war or domestic strife. Dictators clearly diminish economic freedom even in peace time. Austria is a weird one – WW2 a factor, but today is one of the richest in Europe. What’s up with the dire returns compared with neighbours?
So – which countries will stay conflict-free, have a free market, accountable government, low corruption and decent population growth over the next 50 years? Hmm… I’ll probably just stick with a 10% allocation to an EM tracker thanks!
Index tracker reduce risk by spreading your investment over many companies weighted by market capitalisation. Expanding this argument across international borders should further diversify your risk. There may still be some currency risk if the pound strengthens considerably against other key currencies (notably the US dollar as this is 48% of the fund).
The easiest way to do this is to buy a global ETF like the Vanguard All World ETF (VWRL) with a 0.25% expense ratio. This gives you a portfolio of 2623 companies which make up 90% of the global market capitalisation.
I guess those factors are important, but don’t forget technology. Estimates vary on how important it is for growth which seems to be from anywhere between 3-10%, and playing an increasingly important role for the productivity statistics as well as in industrial strategy and survival.
I guess it will only continue to become more intense as technologies spread across different industrial sectors – ICT to automobiles, robotics to healthcare etc. changing how people work, and opening up new sources of revenue and job creation/dislocation.
On that score, the US seems to be doing pretty good in somehow creating an environment where these kinds of firms can flourish, especially in ICT, aerospace (due to defence), bio-science etc. Given the nature of how these things tend to accumulate and be rooted geographically, I guess the US will continue to see some kind of lead from its ability to tap and exploit technology, countered by any policy mistakes that might occur should the tea party loonies cut all R&D expenditures or education expenses from the federal budget; or immigration policy reduces the opportunities for bright students from India/China etc to go and work there.
As for other countries, who knows – but do they have the conditions to enable people to protect their property rights (i.e. IPR), gain financing for their ventures, or even perform R&D and tech development to a decent level? Most are moving in the direction of the US model and it’s going to be interesting to see how they can work it – e.g. Singapore’s development of this model over the past few years is the most obvious case. Korea’s move now to the frontier is also pretty interesting? Japan had difficulties making the switch over from follower to leader, so will Korea follow or depart from this experience?
Either way I guess what is happening is that we are moving more towards the “great convergence” talked about by Mahbubani rather than one country coming to dominate all others.
I suppose this leads to more emphasis on global trackers as mentioned in the post rather than anything too country specific, unless you can identify countries that are pretty good or at least moving in the right direction towards the US type innovation driven markets.
Regarding New Zealand, I think I remember from Shiller’s charts that their stock market did see quite a bit of volatility and big drops as well. I’m surprised to see the 6% real return – that’s incredible!
The NZ stock market is tiny compared to Australia, UK and US. Even the South African market is much bigger.
A well written and interesting history of New Zealand economy:
New Zealand has a wealth of resources – it had a small gold rush, ample valuable and good quality timber, “green gold” in the form of abundant pastures and a guaranteed customer (UK trading treaties until very recently).
Wikipedia example – in 1955, Britain took 65.3 percent of New Zealand’s exports, declining when the United Kingdom joined the now the European Union in 1973. Share of exports to Britain fell to only 6.2 percent in 2000. With the fall in trade to the UK, NZ has increased trading to Asia and now China is a significant trading partner – in 2008 the two countries signed a comprehensive Free Trade agreement.
I wonder how much the close connections with the UK contributed to NZ and Australia’s fantastic returns in the 1900s?
Interesting article on long-term results of world stock indexes, and nice tables. International diversification could actually result in much poorer performance, if certain political actions like confiscation happened to you as a citizen of a certain country.
For example, the German investor from 1900, who diversified equally between Germany and say USA, might have had his US shares confiscated after the country lost the war in 1918. A sore example is the US company Merck, which was the US subsidiary of German Company Merck KGA, which was essentially taken by the US government from the German company. Any investor who owned Russian shares in 1917, Chinese shares in 1940s, or Polish, Czech, securities in the 1940s lost everything. I don’t think this would happen to investors in US or UK shares today, but I can’t say the same thing about those who buy hot emerging market securities.
Either way, great food for thought 😉
Dividend Growth Investor
Do you have this data for the last 20 years that you could share?