A falling price is supposed to recruit its own buyers, and the mechanism failed here for a simple reason: a yield rising on fiscal supply has no level at which it becomes automatically cheap; it only has a level at which the next holder insists on being paid more. That is why the Nihon Keizai Shimbun opened its account of the turmoil with the adage about not catching a falling knife rather than with the one about value [3], and why anxiety over deteriorating public finances, not a growth surprise, is what the paper put behind the move above 3% on the 1st [1].
The long end is where a rate story turns into an accounting one. Forty-year paper is trading about seven points under par [1], and 25-year bonds printed below 80 yen per 100 of face at one point [7], which is more than a fifth of face value gone [2]. Put a modified duration near 17 on that 25-year bond (my assumption, not the Nikkei's) and 20 points of price decline implies roughly 1.2 percentage points of yield rise since it was written [3]. That arithmetic is why Ryutaro Kimura at BNP Paribas Asset Management says the inquiries keep arriving from insurers and pension funds asking how to limit valuation losses [8]. The Bank of Japan, which used to make that arithmetic somebody else's problem, has slowed the pace of its purchases [9], and the executive at an eastern Japan regional bank who told the Nikkei the move was too abrupt, and that it was unfortunate to be running asset management just now [4], is describing a job whose risk profile changed without any change in his mandate.
The currency leg is the part that travels. Nomura's Yujiro Goto called the American comments a fairly direct intervention in the Bank of Japan's policy management and expects the pressure to continue while the yen sits near 160 to the dollar [13]; the yen reached the 156 range on the 3rd, its first visit in about a month [14], which is around 2.5% of appreciation from Goto's threshold [5]. For a Japanese institution, a domestic 10-year at 3% and a firming yen change the relative appeal of unhedged foreign paper without anyone announcing a decision, and that is the transmission line to watch rather than any single auction.
My view, and it may well be wrong, is that the buyer base repriced rather than the level, because when the domestic holder who used to sit to maturity starts asking how to limit valuation losses [8], 3% reads as a waypoint. The counter-case, or rather the more interesting version of it, is positioning: the 3.027% print gave back roughly a tenth of a percentage point within a day [4], which is what a squeeze around a budget headline looks like, and if the Bank of Japan tightens while the yen keeps firming, the same insurers become buyers at these prices and the 3% handle turns out to have been the top. The thesis should fail on evidence rather than argument, and the evidence would be long-dated auctions clearing with domestic institutions adding duration instead of writing memos about how to account for it.
