Despite very easy and low cost access to global equity markets, international portfolio diversification has not been fully embraced yet. Institutional investors are more inclined to build global diversified equity portfolios compared to retail but both parties are still suffering from the so-called “home bias”.  This phenomenon reflects investors’ large preference for domestic equity holdings and a reluctance to diversify more in international markets.
Up to this day, academics are still discussing the origins of home bias and no consensus has been reached yet. Several reasons have been proposed but a mixture probably goes a long way in explaining the effect:

  1. Direct barriers (capital controls) and higher transaction costs. With financial globalization this should be less of an issue now compared to say ten years ago.
  2. International exchange rate volatility adds another layer of uncertainty which scares a lot of investors. But currency risk should not deter investors from diversifying internationally as studies often show that the benefits outweigh the currency risk. Investors concerned about short-term volatility can hedge but for long-term investors, the so-called risk reduction benefits of hedging rapidly declines according to researchers from the London Business School.* Hedging in the long run is costly and it introduces a new kind of risk: a bet on real interest rates at home versus abroad.
  3. Investor psychology. The so-called “familiarity bias” leads investors into the false belief that a domestic stock is safer just because of its location.
*Credit Suisse Investment Yearbook 2012, Dimson, Marsh and Staunton: Credit Suisse Investment Yearbook 2012

Over the past decade, thanks to investor education, international diversification has found its way into investors’ portfolios. But although the evolution has gone into the right direction, the bias is still very noticeable. The table below gives an overview of the domestic bias values from 2001 to 2012 and shows that – despite the steady decline -  all countries still hold significantly home-biased portfolios.Table 1: Evolution of equity home bias per country*


*Determinants of home bias puzzle in European Countries; International Review of Management and Business Research; March 2014

Home bias is also very noticeable if you examine the average equity portfolio of private banking clients. There is a good reason why several private bankers prefer to stay at home and limit international diversification despite better alternatives. Short term disappointing performance will be more forgivable if it happens in a domestic equity portfolio compared to a global equity portfolio. Does this mean that a wealth manager should completely look away and offer only home biased equity portfolios? No, there is an acceptable solution for this:

  1. Any self respecting wealth manager should point to the potential profits of global diversification and try to offer ‘global solutions’.
  2. However, there’s no point in holding eternal discussions with a prospective client who only wants domestic exposure. Wealth managers should therefore also offer domestic portfolios.

In the end, the ultimate choice should be left to the client but not without first pointing out the benefits of international investing. Let’s go through some of these benefits:


There’s enough of historical evidence that global diversified portfolios should earn a higher return for the same level of risk and take less risk for the same level of return. In the short run anything can happen and both higher and lower volatility levels are possible. In the longer run, more diversification should lead to lower volatility if correlations between global equity markets are low enough.

Due to financial globalization, global equity market correlations have been on the rise since the eighties. The chart below shows the evolution of a century of global equity market correlations until 2007/2008.Figure 1: A century of global equity market correlations


*A century of global equity market correlations; Dennis P. Quinn & Hans-Joachim Voth

As you can easily observe, global correlations ran very high by the end of 2007. High correlations spell lower diversification benefits. So surprisingly, this chart goes straight against our initial argument. However, more subtle observations can be made:

  1. Correlation is never constant. Observe the cyclical, mean-reverting nature of the correlation. Periods of high correlation are almost always followed by low correlation (and vice versa). Except for the last two decades.
  2. The diversification benefits were at its lowest point in over a century during the last decade.
  3. Is it unreasonable to expect this correlation to drop again, like it did in the past? Probably not. Extreme correlations never last forever.

And this is exactly what we have seen over the past 3 years. Global equity market correlations are back at very attractive levels providing plenty of diversification potential.

Figure 2: A more recent look at cross market correlations


*Guide to the markets; 1Q 2015; J.P. Morgan Asset Management

Some investors still believe domestic equity markets never go bust. Or they falsely believe that even if it did occur in the past or to other countries, it will never happen again or at least not in their home country.
We think it is not irrelevant to point out that, at some point in the past century, 4 of the 15 largest stock markets in the world suffered a complete loss of -100 percent return. You read that correctly: a complete wipeout of a domestic diversified equity portfolio is no black swan. The markets are:

  • China
  • Russia
  • Argentina
  • Egypt 

Two other stock markets came very close to a complete collapse: Germany (2X) and Japan

Almost every major country has had a bear market in which share prices have dropped over 80 % and some countries have had drops of over 90%. The Dow Jones fell 89% between 1929 and 1932. And more recently, the Greek stock market fell over 97% between 1999 and 2012.

Pretending that a huge bear market will never happen to your country in the next 30 or 50 years is an adventurous bet given that history is filled with these accidents.  Claiming that you will completely time the next big bear right (it will come at some point, only guessing when is an almost futile exercise) is not a daring bet but a rather funny bet. We wish you good luck with that.

No matter how you slice and dice the data, global diversification makes sense – even if correlations are sky high – if you consider complete loss of capital to be the main risk of equity investing. By diversifying country risk, you diversify:

  • Monetary risk
  • Political risk
  • Economic risk
  • Regulatory risk
  • Inflation risk

If you have ever talked to a financial advisor about (potential) equity investments, there is no doubt in our mind that you have heard the line: “In the long run equity always wins”. Maybe your advisor has even presented you these two popular charts:
Figure 3: Rolling total 10 year annualized nominal US returns (1926-2014)


*Weghsteen chart​

Figure 4: Rolling total 20 year annualized nominal US returns (1926-2014)


*Weghsteen chart

Each point on the first chart shows you what a notional investor would have earned (dividends reinvested) annually over the past 10 years if he had invested in the S&P 500 index. One could easily draw the (false) conclusion that negative 10-year periods are very rare. The second chart shows you the nominal rolling annualized returns for 20-year periods since 1926. As you can spot immediately, there has never been a 20-year period with negative returns. And there we have it! Evidence that if you hold on long enough to your (domestic) stocks, you will always come out ahead?

If you calculate the real (after inflation) annualized return of US stocks between 1900 and now you come up with a number around 6%. So an often used argument is that you can expect to earn around 6% above inflation with stocks if you hold on long enough.

But that argument is wrong in so many ways that you should avoid any advisor that uses this number to his advantage. Here is why:

  • The 6% is a long run average and does not take into account the range of all outcomes. What you should see is the full historical distribution of returns for several rolling periods. An average is just an average. The extremely low return periods are much more interesting as it gives investors an idea of what can happen when things turn really bad.
  • The 6% is based on US equity markets only. Is this number valid for other equity markets as well? To get a true sense of what is realistic, we should look at a global history of equity markets. Maybe the US is just a special case and the 6% is more due to luck. Maybe the US is just a market that happens to have survived for so long thereby skewing the results in a positive way.


  • The 6% is backward looking. There is absolutely no guarantee that because something has produced a 6% real annualized return in the past, it will do so in the future. To get an idea of the future, an investor should try to estimate expected long run real returns. Luckily, there is some consensus academic evidence that valuation multiples of national equity markets can help investors in forming an opinion of what the future might bring. We would like to emphasize the word “some”. Calculating expected real returns is still prone to a high error rate. And actual results will differ from expectations as it should be in a future that is ALWAYS uncertain.

Thanks to the work of professors Dimson, Marsh and Staunton we now have deeper insights into long run global equity performance numbers. Dimson et al looked at 103 years of stock returns for 16 countries each.  Their extraordinary work sheds a light on questions like:

  • Do equities always win in the long run?
  • Are US historical risk premiums realistic?
  • What is a realistic historical risk premium for equities?
  • Should we diversify globally or not?

We strongly recommend interested investors to read all of their work (book, papers and their research in the Credit Suisse Global Investment Returns Yearbooks). We’ll present you a couple of their core findings.

First of all, have a look at figure 5. It shows the total distribution of the annualized real returns on US equities over periods of 10 to 103 years.

Figure 5: Annualized real returns on US equities over periods of 10 – 103 years


*Irrational Optimism; Dimson,Marsh & Staunton; December 2003 Version 17

As we have seen in figure 4 as well, you can easily spot that there has never been a period of 20 years with negative real returns. Sounds good for long term equity holders? But the professors have a strong warning for the enthusiasts:

  • This might be due to survivorship bias. The remarkable success of the US in the last century is typical neither of other countries nor of the future for US stocks.
  • There is a big sampling error here. Between 1900 and 2002 (the period of research) there are about 84 historic rolling periods of 20 year. A lot of these periods overlap with each other.  Without getting into too much statistical details, it is important to know that to provide statistical precision, researchers actually need non-overlapping periods. Between 1900 and 2002 there are only 5 such periods (1900 -1920; 1920 -1940; 1940-1960; 1960-1980; 1980-2000). As Dimson says: “Extrapolating from only 5 non-overlapping periods allows too much potential error. The narrow range of 20-year returns in Figure 5 severely understates the dispersion that might occur in the future.”

What about other countries? Does the distribution of returns for several overlapping rolling periods look the same for other countries? Have a look at Figure 6.

Figure 6: Annualized real returns on Japanese equities over periods of 10 – 103 years


*Irrational Optimism; Dimson,Marsh & Staunton; December 2003 Version 17

Oops? Japanese domestic equity investors have witnessed 40-year periods in which the annualized real return was negative! Only when you look back at intervals of over 50 years, you could say that the real returns on Japanese stocks have been consistently positive.

Figure 7 provides investors with an incredibly interesting look at the history of global equity performance.

Figure 7: Percentiles of the distributions of 20-year annualized real returns 1900-2002


*Irrational Optimism; Dimson,Marsh & Staunton; December 2003 Version 17

The chart shows the distribution of annualized real returns for 20-year periods for 16 countries. It provides evidence on the extent to which stocks have delivered a positive real return in the long run (20 years). The main observations according to the professors are:

  • Aside from the US only 3 other equity markets have never experienced negative 20-year periods.
  • For the majority of countries, stocks did not provide a consistently positive real return over the long run, defined as 20 years. A lot of countries have experienced periods of 20 years of negative real returns.
  • Average 20-year real returns were positive for all markets but the details are in the extremes (the worst periods give a better indication of what investors can expect when things really turn bad).

On a more positive note, it is definitely worth mentioning that the researchers also looked at the historical experience of a notional investor who held a globally diversified portfolio (Wld in figure 7 stands for World). The world portfolio is a 16-country index based on market cap. The main observations here are:

  • The world portfolio nearly met the 20-year test with the worst period only being -0,2%.
  • Global diversification would have lowered the returns for a US based investor but it would have also lowered risk.
  • Global diversification led to a reduction in volatility. This is apparent in Figure 8 which shows the summary statistics for annual real equity returns for all 16 markets between 1900 and 2002. The world portfolio had one the lowest volatility numbers (standard deviation).

Figure 8: Summary Statistics 16 markets; 1900 - 2002

*Irrational Optimism; Dimson,Marsh & Staunton; December 2003 Version 17

So indeed, for the US, the long run average real return is around 6%. But for other countries it varied between 1,8% and 7,4%. For the world index they were 5,4%. That is before fees, costs and taxes. Maybe a more realistic historical after tax, after fees real return would be 3% -4%? But these are all historical numbers. Investors should be interested in forward looking expected real returns.

This brings us to the topic of valuations. Yes, investors should allocate globally but they should also take into account stock market valuation levels. Without going into details (that is food for another blog post) we can say that:

  • There is some historical, international and statistical evidence (both from practitioners as from academics) that stock market valuation levels can help investors to determine their allocation levels to certain regions.
  • It is important to point out that valuation multiples are only helpful for periods over 5 or 7 years and more. There is a relationship between valuation multiples and longer term returns. There is hardly any statistical relationship between valuation multiples and short term returns.


  • Global investing definitely has its benefits. It reduces some risks and it could increase the probability of achieving positive 20-year real returns.
  • Global investing should take into account regional stock market valuations. Tilts should be made towards regions with lower valuations.
  • Global investing can be implemented with index trackers in a smart, efficient and low-cost manner.
  • To further increase the probabilities of long term positive real returns investors could allocate toward proven stock market factor such as
    • Quality small cap stocks
    • Value stocks (dividend, value, deep value)
    • Momentum stocks
    • Low volatility stocks

However, investors should also be aware that the implementation of factors - instead of using market capitalization based investing - leads to higher tracking errors. There will be years in which the benchmark will NOT be beaten. Investors should think hard and long before they want to deviate from market cap based investing. There is no free lunch. Shooting for higher returns introduces risk:

  • More volatility
  • More periods of underperformance versus the benchmark

If your risk tolerance doesn’t allow for higher risk, you should stick to market capitalization based  diversification. What is the point of trying to beat the market, if you are going to abandon your plan because of short term underperformance?

We’ll look deeper into the opportunities and the cyclical nature of factor investing in a next blog post.

Hans Heytens