There are 2 problems with the above analysis, and why I think Vanguard skews higher.jalbert wrote: ↑Sun Sep 30, 2018 10:26 pmHere is a white paper describing the results of Vanguard’s research on this. See figure 3 where it shows that 30-40% non-US stocks minimized variance in their sample.
https://investor.vanguard.com/investing ... -investing
I think the difference in variance between 30% and 40% non-US stocks in their data is not statistically significant. Vanguard’s overall recommendation in some articles is 20-40% of equities in non-US stocks, but in others they say at least 40% stocks (which their own data does not support, i.e. going higher than 40%).
First, historical analysis is notorious. It is sensitive to the periods being chosen. You have to assume constant volatility, which is not historical true. etc. Not saying it isn't useful, but the results needs to be taken as informative, not as gospel. Second, it assumes that the future will be the same as the past, which has not true. Things change. You need to ask yourself why 30% was the magical number, what things are changing, and what it should be.
Now, to why I think Vanguard has the correct number. The most efficient portfolio should be the market portfolio, if you assume markets are efficient. Over the past 20 years international markets have become much more integrated with the world economy, and hence becoming more efficient. I know I have been banging on about "efficient" but this is the core theoretical concept here, and the market have change.
I would even argue that international companies have done about as well as US companies. The past 10 years have been very kind to highly leverage, large cap, growth, and tech companies. The US is overweighted in these factors. Going back to the past is not the future, we know factors tend to go in cycles.