IR September 2017 Long•Term Asset Return Study. The Next Financial Crisis have highlighted in previous versions of this note - as well as in the mean reversion section of this report - we can't help thinking that we're setting ourselves up for a return to a few negative real return decades ahead in bonds as we venture out towards 2050. International Returns Fixed income is the asset class for which we have the longest dated data series globally. There is definitely a survivor bias in fixed income though. Although the majority of the analysed countries with data back to 1900 in our study have provided positive real returns over this period there have been some notable exceptions with France (-1.2% p.a.), Italy (-1.9% p.a.) and Japan (-0.6% p.a.) all seeing negative real returns. Germany would be the worst if we had reliable data through the hyperinflation period in the 1920s. So this shows that negative real returns in bonds are a real possibility over even very long periods of time. For equities we only really have comprehensive returns data for a critical mass of countries post WWII and if we look at returns over the last 50 years most developed markets see real annualised returns between +5-6% p.a. The only real notable laggard has been Italy (+1.7% p.a.), although Canada, Japan and Spain have all provided annualised real returns of less than +5%. Since 1980, the period we have previously identified as being the start of a secular global bull market, virtually every country has a higher return for equities and bonds than their long-term average. A notable exception has been Japan as it obviously went through its demographic boom and bust earlier than others. Since the Euro was introduced in 1999, there is little doubt that equity returns in Europe have been disappointing. However this period did coincide with the global equity market bubble so returns are best compared with the US and UK (both +3.5% p.a. real adjusted) for context. Germ