Speculative yen shorts have piled up to a record. And yet the options market is pricing the summer as if almost nothing will happen. The two do not belong in the same world.
The latest CFTC data, for positions as of 2 June and released on 5 June, put the net yen short held by leveraged funds (futures and options combined) at 105,136 contracts, a further 18,887 deeper on the week. Each contract is ¥12.5m, so the notional yen sold runs to roughly ¥1.3tn. Open interest rose by 89,327 contracts to 559,092. Asset managers are short as well, by 62,814 net: the crowd is not one desk. The powder is close to full.
By the standards of recent carry cycles this is an extreme. The August 2024 unwind, when a built-up speculative yen short was forced to cover and dollar-yen fell from about 161 to 142 in a matter of days, is the template for what that covering looks like. Positioning today has rebuilt to a comparable extreme.
Yet the world the options price is calm. At-the-money implied volatility, the market’s expected swing struck near spot, runs 7.2% on the front tenor (the 9 July expiry, which spans the BOJ meeting), 7.7% at two months, 7.9% at three, 8.4% at six and 8.7% at a year: low, with the front the cheapest of the lot and the curve sloping up only gently. In ordinary conditions dollar-yen runs 8-10%; under stress, 15-25%; in the August 2024 unwind it jumped to 25-30%. The current low-7s are the floor of the normal range (Sentry Derivatives, as of 9 June).
The Bank of Japan is expected to lift its policy rate by a quarter-point at the meeting on 15-16 June; the overnight-index-swap market puts the odds at about 93% (TOTAN ICAP, 8 June). Being all but priced is what stops the hike from being the spark. A rate rise that arrives on schedule moves the yen little. The absence of an event premium in July-dated vol, which straddles the meeting, says the same thing: the market reads 15-16 June as a non-event.
Under the yen sits the Ministry of Finance. It spent ¥11.7tn defending 160 between late April and late May, and reserves fell by $77bn that month to $1.306tn. Around 160 a deep-pocketed buyer appears that does not care about price. That works like a put written under the yen: it caps the downside on a long-yen view. The ammunition, though, is finite.
Here the structure turns reflexive. The MOF funds its yen buying by selling US Treasuries, and selling them in disorder pushes yields up. The one thing the Treasury secretary, Mr Bessent, least wants is a rise in the Treasury yields that feed mortgage rates. So Washington’s policy book points one way: back the BOJ’s tightening, let a stronger yen unwind the carry, pull Treasury yields down and feed housing. The plumbing to fund intervention without dumping Treasuries on the market (central-bank repo lines among it) exists. That last step is a reading, not a confirmed plan; the coordination that would prove it does not show up in public data.
If the reading holds, the policy tailwind has a sell-by date. A weaker dollar-yen takes time to pull Treasury yields down, longer still to pass through mortgage rates into housing and then into how voters feel. To reach the November midterms the fall in yields would need to be in place by late summer at the latest. Work backwards and the incentive to encourage the unwind is strongest from June to August. The driver that matters most, the wide US-Japan rate spread, lies outwith the BOJ’s control, which is why timing collects on Tokyo’s two meetings rather than on Washington’s.
An early move into yen strength in June or July, even a sharp one, is therefore no setback for Washington; it merely pulls forward the move it wants. The one condition is that it not spill into a disorderly jump in Treasury yields – accelerate, but do not detonate. The means to fund intervention without dumping Treasuries, noted above, reads as a hedge against exactly that.
The MOF’s own weekly securities-flow data, released on 4 June, carries an early hint of the same turn. In the week to 30 May, Japanese residents were net sellers of foreign long-term debt for the first time in five weeks (−¥185bn) and net sellers of foreign equities for a second week (−¥1.07tn) – a trimming of overseas assets. Non-residents, meanwhile, sold Japanese equities for the first time in nine weeks (−¥491bn) while buying ¥1.25tn of Japanese government bonds, money drawn to JGBs now that normalisation has put a yield on them. One week settles nothing, but the direction fits.
And here is the heart of it. The options price a June-and-done world: a peak of risk around the meeting, calm thereafter. The risk reversal (the gap in implied vol between equally probable bets on yen strength and on yen weakness) leans towards yen strength at every tenor, but the lean is gentle and fades the further out you look, from about −1.1% on the front to −0.3% at six months and −0.1% at a year. The yen-weak tail, too (a topside break of 160, the MOF line giving way), carries almost no premium near the money. The market prices neither tail with conviction: its read is that if covering comes it comes near-term, and the summer is quiet.
The structure says the opposite. The record short, the swelling open interest and the window in which policy most wants the unwind all cluster in the summer. Low, flat vol on top of a record short and rising open interest is the textbook short-gamma, pre-crisis setup. Crises tend to come out of calm. In August 2024, too, vol was low right up to the break.
In fairness, the market may be right. A world in which the authorities together engineer an orderly bleed, and the calm proves warranted, is a real one. The quarrel with the market is about violence, not direction: the authorities want the unwind too. What is in question is whether it arrives as a disorderly snap (our reading) or a smooth bleed (the market’s). A large powder charge does nothing until something pulls the trigger. What the current vol surface states is a divergence (if our reading is right, summer volatility is priced cheap), not an answer.
Watch character, not the number. Not a simple rise in the VIX, but whether the back-month risk reversal steepens towards yen strength, whether credit spreads (the BofA high-yield OAS) widen alongside, and whether dollar-yen, at some level of stress, flips from up (yen weak) to down (yen strong) and accelerates. When those line up, the sleeping market wakes. For now only the front, with its thin lean, has one eye half open on the June meeting.
This piece describes a direction, a structure and a gap in how the market is priced. What to hold, when and how much is the reader’s call. One thing is firm: a record one-way bet is hard to unwind quietly, however much the authorities wish it.