China: a change in FX reaction function

* The Fed minutes were largely as expected, in Macro Man's view.  Yes, they said it was a close call whether they moved in September or not, but we kind of knew that already from subsequent speeches and the three dissents.  That some repressionistas favoured quasi-ZIRP, perhaps forever, was also not particularly surprising given some of their speeches.  That bonds reversed earlier weakness over the course of the afternoon was not altogether surprising; on the continuation chart, the long bond has reached another little "corridor of uncertainty" that has marked previous price extremes.

    
* There was a Reuters sources story yesterday suggesting that the ECB will discuss measures such as deviation from the capital key, buying bonds yielding below the depo rate, and potentially increasing the amount of any given issue eligible for purchase at its upcoming meeting.  Of course, policy won't be changed, but these sorts of discussions may lay important groundwork for any policy shifts announced in December.   Bunds remain close to their lows since the September ECB, having sold off nearly three points since the "taper" story hit the wires.  If anything, yesterday's report may represent a small push in the opposite direction.    Macro Man's Euribor spread trade has worked better than he possibly could have imagined; prudence dictates taking something off the table, which he's done this week.

* The UK is a divided nation, and recent headlines will have left half of the country fuming while the other half gives a wry chuckle.  Yes, the word that mega-retailer Tesco has pulled Marmite (a rancid brown slurry left over from beer-brewing that some people inexplicably put on toast), among other Unilever products, from its shelves has made front page news.  In seriousness, Unilever's decision to raise prices on a broad range of products is just the first shot across the inflationary bow in terms of fallout from sterling weakness; this, and ongoing public angst over sterling's decline should certainly raise inflation expectations.



*  Finally, a bit more on China, where moves in the currency continue to draw attention.  Readers may recall that Macro Man broke down the change in Chinese FX policy in early June, observing that this year the CNY largely ignored rallies in the DXY while participating fully in the down days.   During this period, of course, the RMB depreciated steadily against the reference basket.  However, as Macro Man pointed out the other day, the RMB has been broadly steady against the basket in recent times.  Has anything changed?

By way of reminder, let's look at the state of play as of June 10, when your author published his original post:

As you can see, the RMB displayed a high correlation and relatively largely beta when the DXY declined the day before, but little correlation or beta on days the dollar strengthened.  However, you can clearly observe the change in trend over the last three months, when the RMB has gone more or less nowhere versus the basket.  Is this a reflection of the DXY, or has SAFE changed its reaction function in setting the fix?   Turns out it's a little bit of both.

The RMB is still more responsive to dollar weakness than dollar strength, both in terms of correlation and beta.   However, both the correlation and beta of the RMB to dollar weakness over the last three months is substantially less than that observed earlier this year, in 2015, or any of the time periods studied in the post linked above (back to 2010.)  Now granted, this is a relatively small sample size, so it could just be noise or a temporary phenomenon.   Still, given how pronounced the relationship was to dollar weakness earlier this year, it seems unlikely to be a coincidence.   Rather, it seems likely to be more of an overt attempt to reduce volatility and increase stability.  This in turn presumably stems from a desire to limit the type of headlines that might encourage capital flight.

To put the year in perspective, here are the numbers for 2016 as a whole:

As you can see, the overall theme of a skewed response function fits with what we have observed: namely a fairly sharp (by historical standards) depreciation of the RMB versus its basket.  That the reduction in correlation, beta, and skew has accompanied a sideways move in the RMB versus the basket may seem tautological, but it's not; instead, it reveals that for the last few months at least, the reaction function of the authorities has changed from one of active depreciation to one of stability.

Bear that in mind the next time you hear a frenzied report that USD/CNY has fixed at a new high!