One of the reasons analysts have used to explain what they thought were "too low" JGB yields is that Japan was running persistent current account surpluses. In fact, there was a theory floating around a year or so ago (before this blog was launched) which argued the Japanese current account was moving toward a deficit and so it would be "Game Over!" for the JGB market when that happened. (Since most foreigners have no appetite to buy JGB's.) Japan has now been running a current account deficit, but the 10-year JGB is still stuck at a level that even I think is too low at 0.67%.
The Japan Times discusses the recent record-wide Japanese current account deficit. (The current account is a "wider" notion of a trade balance, as it includes cross-country income flows.) It has hit an impressive-sounding ¥592.8 in November 2013. When you translate that into USD, that is a less ominous sounding $6 billion. For an individual, $6 billion is a lot of money, but that is a manageable amount when scaled by the size of national economies.
By way of comparison, the U.S. Treasury just released the October TIC data (link to latest release). Holdings of U.S. Long-Term Securities by Japanese entities (government plus private sector) was $1,818 billion, of which $1,113 billion was Treasury securities. (And that does not count things like Japanese-owned factories.) This means that just their U.S. securities holdings can finance the present-sized current account deficit for 25 years. And that is just U.S. holdings; the Japanese have snapped up "high yielding" bonds all over the world (which is all of them, when compared to JGB's). Foreigners need not be a net purchaser of any Japanese assets for a very long time in order for the accounts to balance.
Developed countries typically have an equity market capitalization sized at about 100% of GDP, and debt market instruments of over 100% of GDP. Portfolio managers move their holdings back and forth across these markets at a hyperactive rate. A trade imbalance is typically around 3%-5% of GDP, and is a tiny residual relative to these portfolio flows.This is why the best course of action for a fixed income analyst in the developed markets is to ignore trade data.
(Developing economies have less integrated financial markets, and less wealth is held in the form of liquid securities. As a result, these economies blow up regularly due to trade and financial flow imbalances. However, opening their financial markets will probably be dangerous. The developed countries built up their financial markets and wealth behind the walls of capital controls during the Bretton Woods era. How developing countries should manage international financial flows is a topic I know little about.)
I do not have a formal stock-flow consistent model to deal with this cross-country analysis, partly because attempting to model the reaction function of investors is a hazy topic. But it is clear that the portfolio shift effects will dominate the impact of trade imbalances given the size of investor portfolios. In fact, you could probably drop trade flows from your model and it will work just as well.
Returning to a related question of the level of the yen, I do not see any particular risks of out-of-control weakness due to investor over-reaction to the trade data. Japanese corporations borrow in yen, and pay most of their salaries in yen as well. If the yen drops by 1%, Japanese multinational profits rise by more than 1% (this is "operational leverage"), not even taking into account the possibility that the Japanese firms could gain market share. This means that these stocks are more valuable to foreign value investors. The yen can fall for any number of reasons, but it will just create an increasing constituency of buyers of Japanese assets, and so the dynamics will eventually shift the other way.