The book is a nice overview of a selection of well-known asset allocation models, such as the '60/40' (50% US S&P500 and 40% US Government Bonds), 'Risk Parity', 'All Seasons', 'Permanent', 'Rob Arnott', 'Marc Faber', and 'Warren Buffet'.
The numerous graphs provide a ready comparison of the relative performance of the various asset allocation models, often including 'back-calculated' results for times before the particular allocation model came into use. The tabulated results also include a useful 'maximum draw-down' percentage, which I feel is a better yardstick for determining if a particular asset allocation model would suit ones personal risk tolerance than the more frequently quoted standard deviation. After all, it isn't likely to be the "average" risk (variability) that will make you loose your nerve and switch investments, but then "worst case" situation that suddenly arises and induces you to sell out of the market at the bottom and crystallize losses.
The graphs overall show that as a general rule any 'diversified' asset allocation can provide a particular rate of average return for less volatility or maximum draw-down than an undiversified portfolio. Perhaps the only example of a 'free lunch' available in investing. The graphs also appear to show that in the long term (assuming you remain invested over the long term) almost any asset allocation that includes growth assets (real estate, shares) will outperform cash by a considerable margin. However, the use of log-linear axes in the graphs, although useful for showing the consistency of long term growth, tends to make it appear that many asset allocation models have almost identical long term performance. In reality, a fairly minor difference in the 'slope' of a log-linear plot can mean a huge difference in the final dollar value of an investment. Just compare the end result of investing $10,000 for 25 years at an average rate of return of 11% vs. 10% -- while the slopes of a log-linear plot would look very similar, the end result would differ by 25% in dollar terms.
On the other hand, while a small difference in average annual return would have a large impact on the final value of your investment portfolio, the graphs also show that the different asset allocation models have performed very differently in different economic conditions. So the 'ten year average' return for a particular asset allocation will vary a lot, depending on what date range you look at. Given we don't have a crystal ball to foresee economic conditions over the next 10, 20 or more years, it becomes a somewhat moot point whether model 'A' or model 'B' has had superior performance over any historical time period. One therefor comes back to reflecting upon one's risk tolerance -- do you want an asset allocation that has slightly lower average return, but has less 'risk' (volatility) and a much lower 'maximum draw down', or do you opt for an asset allocation that might produce significantly superior returns over the long term, but *might* perform rather badly over your personal investment timeframe, even if that is 30 or 50 years?
Overall, the book is a quick and enjoyable read and provides a ready reference to the performance and risk characteristics of various asset allocation models over the period 1973-2013, and for each decade (70s, 80s, 90s, etc.). And well worth the US$3 price for the eBook version compared to the extra time it would require to gather and summarize this performance data for yourself.
You can get the Kindle version from Amazon here: Global Asset Allocation: A Survey of the World's Top Asset Allocation Strategies
Subscribe to Enough Wealth. Copyright 2006-2016