packer16 wrote:How about adjusting a non-US index by the US dollar index over time or just comparing the non-US index in local currency versus a dollar denominated index.
[...] As AlohaJoe said, trying to determine if the USD was weakening or strengthening at times implies to compare to a basket of other currencies, and I am not sure how to weigh them. Maybe we could do a simple average of the annual exchange rates (an equal-weight approach!), and this might be good enough to get an idea?
I gave it a quick try. For every year between 1970 and 2016, I computed an equal-weighted average of the exchange rates between US$ and the local currency of the 18 countries from MSCI. Then plotted the corresponding values. And this was going up and down like crazy, and frankly I couldn't extract anything useful out of it, it was way too sensitive to the start date.
Then I had another idea. I have the (nominal) annual returns of each of the 18 countries tracked by MSCI since 1970, expressed in either USD or local currency. Why not define an equal-weighted portfolio made of those 18 data series (including the US), and check the USD version of it, or the local currency (for each of the 18 countries) version of it. One might argue the latter better reflects the 'intrinsic' performance of the various local economies (I guess ?!).
Then let's add MSCI USA (expressed in USD) as a third portfolio. The following chart is a telltale chart, with the reference being the S&P 500. In other words, this tracks the growth of each of those 3 portfolios relative
to the growth of the S&P 500. Such telltale charts are great at giving us a good idea of what happens over time.
What can we see? Well, first, MSCI USA and S&P 500 are well in sync (unsurprisingly). Next, this equal-weighted 'global' portfolio would have done quite well. More interestingly, the premium of this equal-weighted portfolio when expressed in USD is a tad better than when expressed in local currencies, but this doesn't change the premium very much. So yes, when the USD depreciates over time, the returns expressed in USD are getting a boost. And as we can see on the chart, the boost can go both ways. Actually, if we had started in 1974, the two lines would have overlapped much more.
I am not sure we can draw that much of a conclusion here, except that currency risk can certainly play games with your portfolio... And as Larry pointed out, this goes both ways.