I am a big believer in foreign diversification for stocks, bonds and cash. Many clients have foreign bonds with a CEF that is mostly Australia and an individual issue from Norway. As time goes on I can see adding individual issues from other countries. These are difficult to access for individuals as the minimum at Schwab is $100,000 per trade. I can allocate to clients as small as I want however.
There are several reasons why I believe in doing this. While I think the dollar is generally headed lower, slowly over time there will be periods where it is strong (bonds from other countries is one way to benefit from a weak dollar). The strategy here needs to be one of looking at the long term. Also, the blending of assets from different countries tends to reduce volatility. This especially makes sense when you consider that the middle of the US curve is very low by historical standards. I think it makes sense to expect that the curve will normalize at some point, and when it does I think intermediate and further out maturities will be higher in yield and shorter maturities will be a touch lower.
This normalization, if it happens, means bond prices would go down. As a rule of thumb, a 1% increase in rates works out to about an 8% decline in price -- of course shorter maturities work differently.
Another aspect to this is that I think other countries are better off economically which makes for a compelling investment destination for stocks, bonds and cash.
I am far from the gloomiest guy on the future prospects for US capital markets, but I do think that better results will be had over the next decade or so by looking more to foreign. I have been about 35% in foreign and can envision increasing that to 50% (a point I have made quite a few times in the past). My foreign exposure to bonds is quite low currently, but I could see that going to 25% in the future. Of course if things really come unglued in the US I could go heavier with foreign bonds but I don't think it will come to that.
Constructing a foreign bond portfolio is not dramatically different from constructing a foreign stock portfolio. I would have a mix of different countries some with surpluses and some with deficits, countries with different types of economies at different points in their respective economic cycles and varying interest rates.
The big point here is you probably need to learn about this if you have not already done so in order to make your own decision. There is no way, now, to know how important this will become. Having to learn about this on the fly during some sort of unraveling seems less than ideal.
I would expect that accessing this market will get easier as fund providers bring product to the market (at least I expect this to happen). MSCI has created a lot of indices for this space - someone just needs to license them and create the funds.