1. Have Japanese government bonds moved two standard deviations in any period since 1990? (I don’t know…)

2. Would you characterize them as a bubble bound to eventually burst messily? (Note Japanese government debt has tripled to 180% of GDP even while the savings rate has been declining, yet Japan’s government CDS spreads are lower the than US and many other countries!)

I am suggesting that a move so large is ALWAYS a bubble irrespective of the underlying fundamentals for why it happened. An example would be the upturn in commodities this past year. I happen to think we are seeing peak oil. But did peak oil justify $147/barrel?

So, I am making an a priori assumption without having tested the fundamentals as to inflation expectations, etc, etc.

But, the evidence does point toward bubble: what are the yields on other government bonds? What are the yields on corporate bonds? Why are treasurys outyielding TIPS? Why have treasurys returned over 20% this year? Why the discrepancies? Is it inflation? Is it risk aversion?

When you have to start asking yourself so many questions before you get into an asset class, you know it’s not cheap. The reason I make an a priori assumption is to avoid investing in manias and taking a loss. It’s just too good to be true.

The recent extreme gains in treasuries have happened in a very short time period that coincided exactly with CPI dramatically falling below expectations. My read on the situation (speculative only, so possibly very wrong) was that the majority of the move was large investors (pension funds, endowments, etc) realizing that the fundamentals weren’t what they thought — i.e., the threat is deflation not inflation. I say majority — of course there were some investors changing performance rather than perception of fundamentals.

As an example of another bubble, the NASDAQ bubble didn’t spike in quite as short a period as a few weeks, nor did it coincide with a change in any fundamental measures of the economy (e.g., CPI) that people pay the most attention to.

My simple explanation is something I tend to post on at some point: I call it the two-standard deviations rule. Basically, when any tradeable assets reaches a point two standard deviations above trend a self-fulfilling mania seems to occur. It’s as if a tipping point is reached where the “random walk” is no longer in play. It is what I would call “non-independent errors.

In plain English: when an asset goes up enough, everyone thinks it’s a sure thing and piles in. This bubble-like group think has already started to happen in treasurys.

While a bubble can last for year: look at Japanese real estate or u.s. housing: it is still a bubble from the point when these non-independent errors start seeping in. The whole thing could collapse at any time — and, therefore, these are asset class to trade in at one’s own peril.

Perhaps your point is that prices of other assets already discount this outcome — but again, how can you be so certain in this assessment?