Our Insights

About personal finance, investments and markets.
February 27, 2017

Looking backward to look forward: 117 years of investment returns

Why do institutional investors like pension funds all have similar investment strategies? Because they have access to the same historical data and they can see what works and what doesn’t. But those data are no secret. Credit Suisse has just published its latest “Global Investment Returns Yearbook” which provides an analysis of investment returns stretching back 117 years. This research paper helps to give context and offers invaluable perspectives. We highlight here our three key take-aways. Please get in touch If you want to understand how we use the data to structure our clients‘ assets to help them reach their financial goals.

1. Long-run returns of the major asset classes

What is the long-term perfomance of equities, bonds and t-bills? The authors Elroy Dimson, Paul Marsh and Mike Staunton have crunched the numbers every year since publishing their highly recommended book „Triumph of the Optimists“ since 2002. The following chart shows the cumulative total returns from 1900 to 2016. Equities returned 9.5% per year, versus 4.9% on bonds and 3.7% on t-bills.

Chart 1

On an inflation-adjusted basis, the superior equity returns become even more obvious. While t-bills struggled to achieve positive real returns (+0.8%), bonds increased by 2.0% and only equities offered a path to material real wealth growth (+6.4%). The data include several major wars as well as the financial crashes of 1929, 1987 and 2008 – however, events that were traumatic at the time now just appear as setbacks within a longer-term secular rise.

Chart 2

2. Bonds might not be as safe as you think

Why is a longer perspective so important? Bonds are a prime example. Every investor has probably heard the pitch that “equities are for growth and bonds are for safety”. The financial industry then typically quotes bond returns of the last few decades, which do look quite attractive. From 1982 to 2016, global government bonds delivered real returns of 6.9% without much volatility. But it would be a serious mistake to project these performance data into the future! The performance data from 1982 for bonds are potentially highly misleading because they only capture a certain very bond-friendly environment, where

  • the starting yield was in the double digits,
  • inflation was declining and
  • interest rates too

Who can guarantee that the same will be true in the next few decades? Given that the starting yield for many government bonds like the German Bund is virtually zero, the future return until maturity will be zero – it is a mathematical certainty. At that level, bonds offer only return-free risk.

To provide an overview over alternative scenarios, here are some periods which have seen bond holders nearly getting wiped out. It is important to realize that bonds are not always safe investments.

PeriodRegionTotal loss (real)Comment

A data set of this historical breadth offers investors a reminder that bond prices used to go down as well as up. Since 1900 they have had similar enough real drawdowns and real volatility as stocks with 3.3% lower returns – long-term investors loading up on government bonds today need to understand how unfavorable the risk-reward profile is.

3. Survivorship bias and the importance of the rule of law

Did you realize that most historical performance data are based on the US? Well, it is not just because those data are easier to get hold of, it is also because the US data look so much better than for most other countries! There is a clear case of survivorship bias – investors in some countries got lucky but others suffered financial disaster or dreadful returns. After all, Russian and Chinese equity investors got completely wiped out following the communist revolution. For most of Europe, WW1 and WW2 destroyed a lot of productive capacity and hyperinflation devalued most fixed income assets. What we are trying to say here is that it is dangerous to generalize from US asset returns since they exhibit ‘success bias’.

Chart 3

But all is not lost. If you compare the equity returns of most large European countries with the US from 1967 to 2016, you get nearly identical real-returns.

  • US 5.9%
  • Germany 5.8%
  • UK 6.9%
  • France 6.0%

What this shows, in our opinion, is how important political stability and rule of law are. If you remove either, equity returns are likely to suffer. The threat to US institutions rather than any minor change in the GDP growth rate is why we are really worried about Trump. Our advice: When EM politics is everywhere, buy the assets priced for it. More on this in our next blog post.