Yen

I follow the movement of Yen for a few years. Current level is 118.84/$. I think almost all US investment bankers and many academia now agree that Yen is undervalued. As far as I know, the reason for the current weak yen is driven by the domestic Japanese capital outflow.

Over the past decade, the central bank kept the interest rate at 0%. The liquidity in the market is huge. With the postal reform, money is able (or at least expected to be able) to move abroad. We are talking about trillions of Yen. Of course, if I am a Japanese portfolio manager, I will do it now, expecting Yen will depreciate in the future. This is in contrast to US bankers' view. Of course, Yen is depreciating currently, meaning that the outflow force from Japan is higher than the inflow force from non-Japan. The question is will this force switch?

I looked at the bond indices for a set of countries (Australia, Japan, Switzerland, South Korea, South Africa, Brazil, US, Euro (average of Germany, France and Italy), Mexico, UK, Poland, Chile and Indonesia). I calculated the rate of return and its correlation with the Japanese bond market index. I also adjust the FX exposure return respect to Yen. I separate the horizon into three periods (2000-2001, 2002-2004 and 2005-2006 Nov).

Here is the thought experiment: when Japanese investors decide to invest in the US bond, they form their expectations on the future return, based on current and past information such as exchange rate, bond correlation and hence overall returns. For example, when the bond return is realized in 2002, the decision was made in 2001 (for example). So, available information is 2001 exchange rate and 2001 bond correlation and returns. Exchange rate impact will be realized in 2001 but actual return will be realized in 2002.

I first assume Japanese investors have adaptive expectation, adapted from the past two years. What I found is that if their decision was based on past two years' experience, they were wrong. From 2000 to 2001, from Japanese investors' prospective, investing in US bond market was a better deal than investing in the European bond market both in terms of diversification (correlation) and returns. But we observed that from 2001, Yen was appreciating against the $, but slightly depreciating against the Euro. This implies money flow from Japan to Europe, rather than the US. Indeed, it turned out that the return between 2002 and 2004 was higher in European bond market (13%) than in the US bond market (2.2%), although the risk was also higher investing in Europe (higher correlation) than the US (0.83 vs 0.58). This suggests investors are forward looking.

I did the same exercise from 2005 to 2006. If investors are forward looking and if realized return in 2005-2006 is what they expected, it is hard to know in advance where to put their money. Between 2005 and 2006, Return in Europe is slightly higher than the US (8.5% vs 7.8%). The respectively correlation are 0.59 and 0.23. But again, we observed Yen appreciated against the $ between 2003 and 2004 more than against the Euro, meaning money flow to the Europe more than to US.

If the capital flow story is true, Japanese investors care more on return than risk. 0.7% return differential is enough to compensate for the 0.36 correlation differential. The "excess" savings in Japan are really excess. Search for return is their major goal.

Looking forward, interest rates across the regions are the key variables to monitor.