The weakness in Asian manufacturing output is also prominent in Japan. This week the government revised down the latest figures on IP, and it now appears that the drop in Q3 production may meet or exceed the 5.9% decline in Q2.
Net trade appears to have exerted a substantial drag on Q3 output alongside an apparent drop in capex. An adjustment downward in the pace of inventory building likely also was a drag on activity.
As with the United States, our expectation is that Japanese companies are nearing their objective for inventory growth, paving the way for a pickup in production this quarter. This is being further supported by signs of a pickup in consumer spending, as underscored by this week's stronger than-expected reading on personal outlays for August.
Up-coming reports on September international trade (October 21) and especially IP (October 29) will be important for gauging progress. The outcome of the BoJ meeting on October also hinges on these reports. We maintain our call for renewed easing although this appears to be a close one.
JPY was the star performer through Q3, rising against all G10 currencies. However, the evidence now suggests JPY is being propped up by increasingly skewed positioning and for the first time since the beginning of the Abe/Kuroda era, FX managers appear to be short USD/JPY.
On the face of it, this is compelling evidence that FX participants have a large concentration of risk in short USD/JPY positions. net short JPY positions, having accounted for 70-80% of open interest for most of the last three years have slumped to just 10-20% in recent weeks. Similarly, having been consistently bid for calls, USD/JPY risk reversals have flipped to be bid for puts across the entire vol curve.


BTC Flat at $89,300 Despite $1.02B ETF Exodus — Buy the Dip Toward $107K?
Nasdaq Proposes Fast-Track Rule to Accelerate Index Inclusion for Major New Listings
FxWirePro- Major Crypto levels and bias summary
JPMorgan Lifts Gold Price Forecast to $6,300 by End-2026 on Strong Central Bank and Investor Demand 



