When the yen suddenly surged against damn near everything during one of the more dramatic “flash” events in recent memory just after 9:30 AM in Sydney, it wasn’t too difficult to make a list of possible contributors.
The yen was already on the front foot amid the generalized risk-off mood across global markets and was tracking the ongoing rally in U.S. Treasurys pretty closely. Well, 10-year U.S. yields dove to a fresh 11-month low following the Apple guidance cut.
That probably set the stage.
More importantly, though, Japan was out on holiday and the moves unfolded during the fabled “witching hour” during which similar “anomalies” have manifested themselves in the past. Here’s one example from January 11, 2016, when the rand suddenly plunged 9%(this isn’t the most famous example and indeed, that’s why we highlighted it – to remind you that it happens more often than a lot of folks think):
(Bloomberg)
The moves in the yen were apparently sparked by Japanese retail abandoning the lira and the Aussie – or at least according to most accounts. Algos quickly picked up the moves and once USDJPY plunged through key levels, short covering kicked in, the flow was overwhelming and the rest is history. Or something.
Read more
Another ‘OMG’ Moment: Yen Surges In Incredible Bout Of FX Madness
We documented this in real time in the linked post above, but it’s worth taking an updated look at the charts. Here’s USDJPY and AUDJPY, with the move in the latter serving as the focal point for many postmortems on Thursday.
(Bloomberg)
One-week implied vol. on Aussie-yen exploded, as did one-month dollar-yen implied.
(Bloomberg)
Positioning likely exacerbated the “problem”.
(Bloomberg)
The lira was on the frontlines as well, plunging against the yen and the dollar in a repeat of some of the more manic episodes from July and August. This is astonishing:
(Bloomberg)
For good measure, here’s sterling and also EURJPY:
(Bloomberg)
“The ‘flash crash’ in USDJPY during Japan’s holiday period has highlighted, in dramatic manner, the significance of the domestic bid for yen crosses”, BofAML wrote in a Thursday note, adding that “the absence of Japanese demand coupled with poor early year liquidity led to [the] collapse.”
(BofAML)
Obviously, this serves to underscore concerns about fragility and marks a rather inauspicious start to a year that many hoped would be less volatile than 2018. This is the “new normal” – so to speak. Abnormality is now the norm.
That said, one of the most underappreciated aspects of these types of events is the signaling effect. Sure, it’s highly disconcerting when something like this happens, but what’s even more pernicious is the extent to which, depending on what assets are involved, the dislocations have the potential to send “false” messages to markets.
How does one, for instance, look at where the yen and lira are trading on Thursday and decide what exactly is in the price? There are surely remnants of the flash moves, but how do you parse what’s “real” and what are just lingering vestiges of an anomaly? The yen is obviously a haven and the lira a proxy for risk appetite, and now, market participants’ ability to use those currencies to get a “read” on sentiment is impaired by the futility of discerning what’s actually in the price after the flash event.
BofAML underscores those points in the note cited above. “Perhaps more importantly from a short-term behavioural perspective, the break below 110 in USD/JPY and critical technical support levels in other yen crosses, may lead to a re-evaluation of the domestic dip buying mentality”, the bank cautions, adding that “the majority of Japanese exporters assume USDJPY at 110 for the current fiscal year, while purchases of foreign bonds by institutional investors have been concentrated above 110.”
The bank goes on to say that “the recent price action may encourage more active hedging and less dip buying over the near term.”
So, when you think about “flash” events, fragility and modern market structure more generally, do note that even after these dislocations partially “correct” themselves, the mark they leave is not thereby erased.
There is one flash crash nobody talks about: the General Collateral rate for UST collateral on December 31 was 5.15%, 290 bps above RRP, the reverse repurchase agreement rate. Yesterday another spike occurred again.
There is something wrong somewhere, better to pay attention carefully.
Can you elaborate further about what this means and it’s ramifications, for for instance for treasury market bond ratings/stability’s? Thanks.