Whenever you talk about JGBs, you can calm hysterical people down by saying this at the outset. For bond markets to well and truly collapse, you need two things: Rising nominal yields and lower real yields. The combination means that bond won’t get cheaper when prices fall. That is the disaster scenario. You need both. A rising nominal yield means nothing in and of itself, rising nominal bond yields when accompanied by higher real yields, brings in money in a sell-off. Once that is out of the way, you can move on to the heart of the matter.
So when CHFEUR blew up, it resulted in some interesting action in Japan. One, Mrs. Watanabe blew up my relative valuation model. Further, the effect is persistent and may require a new perspective on the yen and JGBs.
Relative valuation models are just glorified plays on mean reversion. My Uncle Larry would hate hearing it, but mean reversion is how value trading is implemented nowadays. These models are in no way predictive of the future, merely calculating deviations from fair value based on the data inputs. My model was working pretty well until January 17th when the CHFEUR peg blew-up. There were other points of the time series where model fit left something to be desired, but these all centered around QE in early 2013 and Fukushima in early 2011. Then last month you saw a massive, massive move by Mrs. Watanabe back into JGBs—CHFEUR effect. I’m showing you the whole market, not a particular maturity, because I’m not sure of her preferences.
Note that this is pure carry trade unwind back into yen and liquid sovereign yield, and it still hasn’t recovered all the way. You saw no particular inflows into risky debt or stocks. Risk is risk, home currency or not, and there was little taste for it after Schweizerische Nationalbank hosed down the carry trade.
Another interesting thing: What you are seeing is that the carry trade is so big a reversal has dislocated the home currency credit market in that credit risk isn’t catching a bid at all. Households would rather take FX risk in sovereign debt than take credit risk in Japan that much. It is, however, not reasonable to think that sovereign debt is free of credit risk.
If you adjust for term risk differences of the assets, the effect shown is less pronounced but the dislocation would be no less true. Interest rate differentials are becoming the decisive driver of risk in Japanese credit. You can see this in 5Y credit derivatives. Corporate debt got little to no bid in the aftermath. JGBs continue to see buying.
Here you see the extent of the impact of retail money coming back into home currency JGBs reference rates and even seeing de-risking in corporate and bank debt in home currency debt. This is perhaps the effect of dampened risk appettite spilling across asst classes, or the effect of collateral damage to maintenance margin forcing selling in other positions. Either way, you see JGBs catching the bid.
That bid is weird. Which is to say: increasingly negative nominal interest rates are becoming the norm in many parts of the world. This is creating even bigger problems for all the modelers and collateral damage for real money.
You see, long term rates are nothing more than average short-term rates appropriately discounted for the future. The problem is that a bond potentially locks a position into current average discounted rates over a long time period. You are likely levered in the position 2-3x with near-zero funding rates giving you a nice return in the margin. If funding costs ratchet up, there will be a real problem. Rates moving in the other direction will cause problems as well. With negative yields you can be in a situation where a single counterparty ends up paying both legs of an interest rate swap. When that happens, rates derivatives markets will go illiquid and lock up, dislocating markets even further. The downstream effects of this will be severe.